<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Indu Pathak on Medium]]></title>
        <description><![CDATA[Stories by Indu Pathak on Medium]]></description>
        <link>https://medium.com/@pathakindukumari?source=rss-f965785d0636------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*v9oXwpx6D2mtvnPc</url>
            <title>Stories by Indu Pathak on Medium</title>
            <link>https://medium.com/@pathakindukumari?source=rss-f965785d0636------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 01:45:00 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@pathakindukumari/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Why We Upgrade Our Phones Too Soon: 6 Ways To Make Your Tech Last Longer]]></title>
            <link>https://medium.com/@pathakindukumari/why-we-upgrade-our-phones-too-soon-6-ways-to-make-your-tech-last-longer-3fda746a2060?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/3fda746a2060</guid>
            <category><![CDATA[smartphones]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[personal-growth]]></category>
            <category><![CDATA[e-waste]]></category>
            <category><![CDATA[sustainability]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Fri, 13 Nov 2026 12:37:41 GMT</pubDate>
            <atom:updated>2026-11-13T12:37:41.578Z</atom:updated>
            <content:encoded><![CDATA[<h3>My phone was working, but I was sceptical.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1FLRJPxktEX_Un3c-HfmVw.png" /></figure><p>Last year, my phone was getting sluggish — apps were loading just that little bit slower and by lunch I was down to 20% battery. And I kept admiring how good my mate’s new phone was with the camera and feeling like my phone was just not quite good enough.</p><p>Like all of us I went onto the shop page and selected a colour.</p><p>Yet, as I was sitting down to some food with my brother, he asked me “Well, what actually is wrong with it?” Which, it turns out, I couldn’t give a really good answer to. My phone, I could still make calls with. It could take good pictures and all of my apps were still working. It was just old and quite honestly, so was I.</p><p>I closed the tab. 3 months, a new battery and a deep clean-out later, I’m still using it. Little did I realise that by making such a small decision I was going to go down a rabbit hole and discover some fascinating reasons why we upgrade our phones so frequently, and how that can have both a personal and planetary impact. Here are 6 things I found.</p><h3>1. It’s more expensive than you think!</h3><p>We know how much we pay for a new phone, but there’s a much higher price we’re paying in secret.</p><p>In 2022, the world generated approximately 62 million tonnes of e-waste (Global E-waste Monitor 2024, United Nations). Only approximately 22% of it was collected and properly recycled, with the rest ending up in landfills, drawers or dismantled by hand.</p><p>That surprised me. Every time we upgrade our phones early we are adding to that number. And we are also adding to the mining, shipping and energy used to make a new phone.</p><p>In addition to the direct expense, most of the time we’ll purchase accessories: a new case, a screen protector, maybe a new charger or Bluetooth earphones, and, of course, a longer contract. By the time I did the maths, my “little” upgrade had set me back far more than the phone.</p><p>Besides, my little old phone wasn’t going anywhere. It would have been added to the 2 old phones already in my drawer.</p><h4>Try this first</h4><p>Write down what is wrong with your phone. If you have a short list, you probably don’t need a new phone. Repair it or clean it out.</p><h3>2. Often phones aren’t actually old, they’re full!</h3><p>A significant reason for believing our phones are ancient is that they are overcrowded.</p><p>The moment I finally got around to having a look at my phone, it was already almost full. The last time I checked, I had thousands of screenshots, three versions of the same video taken from a wedding and so many apps I hadn’t opened in a year.</p><p>I spent the morning clearing everything out. I backed up all my photos on the cloud and deleted apps I didn’t need. It made a real difference.</p><p>I also had about 15 apps open in the background. So these notifications and location tracking were running for the whole day. It impacts the battery and the memory. If you turn the majority of them off your phone will be way more relaxed and so will you!</p><h4>Simple clean out tips</h4><ul><li>Delete screenshots and duplicate photos monthly</li><li>Delete apps you haven’t used in 3 months</li><li>Clear out old WhatsApp media (this can eat up gigabytes!)</li><li>Restart weekly</li></ul><p>It’s not fun but it’s free!</p><h3>3. Batteries degrade but can be replaced!</h3><p>Typically the very first thing to fail on a smartphone is the battery. It makes your phone feel broken.</p><p>All batteries wear down. But I was amazed how low cost and easy it was to replace! I took my phone into an authorised service place and they replaced my battery in under an hour, for a small proportion of the cost of buying a new phone. It wasn’t long before my ‘dead’ phone was lasting all day again!</p><p>The good news is that new phones in the EU have to hold at least 80% of their battery capacity after 800 full charge cycles, starting from 20 June 2025. That means manufacturers will build batteries that last longer.</p><h4>Take care of your battery!</h4><p>Many phones now offer an option to limit charging to 80%. I have this on and try not to leave it charging overnight in a hot room. Heat is the enemy of batteries!</p><h3>4. Repairs are becoming easier!</h3><p>For years, repairs were a hassle. Parts were difficult to source and some shops gave high prices for repairs meaning a new phone was more attractive.</p><p>However, this is starting to change. The same EU rule means manufacturers must provide key parts such as batteries and screens for a minimum of 7 years after the model has been discontinued. Phones must also provide OS and security updates for a minimum of 5 years after the last unit has been sold.</p><p>Even outside of the EU, these rules often affect the phones we get everywhere. Companies rarely create different models for a specific region.</p><h4>Find a good repair shop!</h4><p>Don’t wait until your screen cracks. Find a reliable repair shop near you. Maybe ask friends or read reviews. If possible, go for an authorised repair shop. Anything to do with the battery is usually safer to go to an authorised shop.</p><h3>5. Security updates are more important!</h3><p>Phones without security updates can be a risk. However, if your phone gets security updates, it can last for years.</p><p>I used to turn off update notifications as they would come up at the worst times. However, I now realise they are important. They help fix any security issues, improve battery management and can make an older phone feel faster.</p><p>When buying your next phone, check how many years updates will be provided. This can make a big difference between your phone lasting 3 years or 6 years!</p><p>Most new features aren’t that impactful. Most launch events show a slightly better camera, brighter screen or new colours. Whilst these are nice to have, they rarely impact how you use your phone. Looking at my phone, I realised I mainly use it for messaging, paying bills, reading and taking photos of my nephew. My 3 year old phone does all of this perfectly!</p><h4>Easy updates!</h4><p>Turn on auto updates and let them run overnight whilst charging. This means you won’t get any notifications and your phone is safe.</p><h3>6. Don’t let your old phone go to waste!</h3><p>There will come a time when your phone will truly be at the end of its life. However, how you get rid of it matters.</p><p>Whilst writing this I found 2 old phones and a tablet in my kitchen drawer! Many of us have a similar collection. These phones still contain precious metals however they need to get to a proper recycling service.</p><h4>Better ways to pass on your old phone</h4><ul><li><strong>Pass it on!</strong> An older phone can be perfect for parents, children or friends who just need to make calls and send messages.</li><li><strong>Trade it in for credit!</strong> Many shops and networks will give you credit for your old devices even if they are cracked!</li><li><strong>Recycle properly!</strong> Take your phone to an official e-waste collection centre. Don’t put it in your bins. Delete your data first.</li></ul><h3>What I’m doing differently</h3><p>I’m not against new tech! I love gadgets as much as anyone. However, there will come a time when I will need a new phone.</p><p>Now, when I am thinking of buying something, I go through a quick list:</p><ul><li>Have I written down what is wrong?</li><li>Have I cleared my storage and background apps?</li><li>Can I get a new battery or screen?</li><li>Is it still receiving security updates?</li><li>Do I really need new features?</li></ul><p>If a new battery or screen would fix it, I won’t buy. It only takes 5 minutes and has stopped me from buying more than one item on a whim.</p><p>I no longer upgrade my phone due to boredom or wanting a nicer camera. I ask what is wrong, attempt to fix it and only buy if I get the answer “this can’t be fixed”. My phone is now 3 years old and works perfectly. My bank account is happier and I feel a little better about that pile of e-waste.</p><p>Often, the most sustainable phone is the one you have in your pocket!</p><p>When was the last time you upgraded your phone? Was it broken or did it just feel old?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3fda746a2060" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Promises We Keep Breaking (And 5 Ways To Start Keeping Them!)]]></title>
            <link>https://medium.com/the-gravity/the-promises-we-keep-breaking-and-5-ways-to-start-keeping-them-0a75a8f69e3e?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/0a75a8f69e3e</guid>
            <category><![CDATA[habits]]></category>
            <category><![CDATA[self-improvement]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[motivation]]></category>
            <category><![CDATA[personal-growth]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Fri, 13 Nov 2026 00:05:27 GMT</pubDate>
            <atom:updated>2026-11-13T00:05:27.477Z</atom:updated>
            <content:encoded><![CDATA[<h4>Big goals are great, but small boring promises are the real game-changers.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cG7yu-L7jtsWIBlzEMr0DA.png" /><figcaption>Image is generated by AI</figcaption></figure><p>Last January I wrote a list of promises on the back of an electric bill. I wanted to wake up at 5.30am, run 3 times a week, read every evening, learn Spanish, drink more water and not look at my phone as soon as I woke up.</p><p>I stuck it on my fridge with a lemon-shaped magnet and by the second week of February it was covered by a timetable and takeaway menu. I had done maybe 2 things, but not very well.</p><p>I’m not alone in this. Most of us are fantastic at making promises to ourselves, but not so great at keeping them. Not because we’re lazy, but mostly because we’ve made promises that are too big, too vague or were made by the best version of us - the one that has had a good nights sleep and a strong cup of coffee.</p><p>This year I made myself promises that were smaller and tried to keep them like I would with a friend. Here is what I learnt…</p><h3>1. The Promise Was Just Too Big</h3><p>Looking back at my list everything on it was huge. Waking up at 5.30am? I was waking up at 7am! Running 3 times a week? I hadn’t run since sports day at school!</p><p>When you make a big promise it feels like you’ve failed if you don’t do it. I would oversleep on Tuesday and by Wednesday I would be thinking ‘this is ruined’ and give up.</p><p>My friend Rekha does this with dieting. Every Monday she sets herself a new goal but by Thursday she has had a biscuit and given up. We find it funny but it really isn’t. If you’re only going to achieve your goal if you’re perfect you’ve basically set yourself up to fail.</p><h4>Make it tiny</h4><p>Now I set myself tiny goals. Goals so small they’re almost embarrassing. Instead of running 3 times a week I might say ‘I will put my trainers on and walk to the end of the road’. Instead of reading every evening ‘I will read one page’.</p><p>Interestingly, once I have my trainers on I usually go further than planned! And if I don’t, I’ve still achieved my goal. Which is more important than I thought.</p><h3>2. 21 Days To Form Habits?</h3><p>You’ve probably heard this before. It takes 21 days to form a habit. I believed this for years. However, by day 22 I would have realised it felt like I was still struggling and assumed I was just rubbish.</p><p>There was a study done on this. In 2009 Phillippa Lally and colleagues at University College London looked at 96 people each trying to create a new habit. Things like having a piece of fruit with their lunch or going for a run before dinner. On average it took 66 days for their behaviour to feel habitual. Some people took 18 days and others took closer to 254 days.</p><p>254 days! Over half a year!</p><p>When I read this I felt strangely relieved. I wasn’t a failure, I was just in the middle of something that takes ages to develop. No one had explained it properly.</p><p>Interestingly the study also found that missing one day didn’t really matter.</p><h4>Give it more time</h4><p>I no longer count to 21. I try to give myself at least 3 months to get into any new habits. If I miss a day, I’ll just do it the next day.</p><h3>3. Be Specific</h3><p>For a long time I would say things like “I’ll try to meditate more,” or “I’ll try to call Mum more often.” This didn’t help as I was using ‘try’ as a crutch. It was never going to help.</p><p>Psychologists have a term for this called an implementation intention. Basically setting yourself a specific time and place to do something. In 2006 Peter Gollwitzer (the psychologist who came up with this idea) did a review of 94 studies with his colleague Paschal Sheeran. People who set themselves a specific intention - e.g. “if this happens, then I’ll do that” were much more likely to complete the task than those just setting a goal.</p><p>Instead of ‘I will try to call my Mum more’ I would say “On Sunday, after I’ve cleared the lunch plates, I’ll call Mum before I sit down.” It doesn’t sound very natural but it works! I will always be having lunch on Sundays so it ties in.</p><p>A few of mine are:</p><ul><li>When my kettle boils in the morning I will write 3 lines in my notebook.</li><li>When I park up at the supermarket I will take 5 slow breaths.</li><li>When I shut my laptop I will write my one important task for tomorrow on a sticky note.</li></ul><p>They aren’t impressive but that’s the point!</p><h3>4. Don’t Wait For Motivation</h3><p>I would wait until I felt motivated. ‘I will start my online course when work is quieter’, ‘I will sort my spare room when I have more energy.’</p><p>Work never got quieter and I never had more energy. My spare room became a box filled room for nearly 2 years.</p><p>Every time I passed my door I felt a little guilty. It’s surprising how much energy I spent not doing something. Eventually I had 4 Saturday mornings, a roll of bin bags and a very dusty exercise bike to give to a neighbour.</p><p>I’ve noticed that I usually get motivated once I start. The first 5 minutes of pretty much anything is dreadful but once I get to 10 minutes in I’m okay. Sometimes I’m even enjoying it by 20 minutes in!</p><h4>5 Minute Rule</h4><p>If I really don’t want to do something I’ll set myself a 5 minute rule. I will do it for 5 minutes and then I can stop. Honestly I only stop 1 in 5 times. The other 4 times I carry on because the worst part was opening the box!</p><h3>5. Track Your Progress</h3><p>This seems simple but it worked best for me.</p><p>I bought a cheap wall calendar from the supermarket. Every day I complete my small promise I put a cross in with my red pen.</p><p>After a couple of weeks I have a chain of red crosses and I really don’t want to break it. On evenings where I’m tired I catch myself thinking about the empty box and I go and do my one page or short walk.</p><p>My nephew noticed my calendar when he came to stay. He asked what all the crosses were for. Once I explained he created his own. He’s only 11 but is much stricter with himself. He even gave me a telling off for cheating with a half cross on a bad day!</p><h4>Don’t worry about missing days</h4><p>Throughout the year I missed some days. A whole week in April because everyone in the house had a cold. 3 days in August because we were out. I used to feel like I was failing but now I just think of them as busy weeks.</p><p>As long as I’m getting crosses in…</p><h3>What My Year Looked Like</h3><p>I didn’t learn Spanish this year. I’m not a 5.30 person and I don’t think I ever will be.</p><p>However, I’ve walked most days since March. I’ve read 11 books (which is a lot for me!) I call my Mum every Sunday and my spare room is now a spare room! With a bed to sleep in.</p><p>It wasn’t one big decision. It was many small decisions made on regular Tuesdays.</p><p>I still slip up some weeks. I still get Monday mornings where I want to rewrite my list. However, I know what to do now. Make them smaller!</p><p>If you have an old list on the back of a bill don’t throw it away. Pick one thing you want to do. Make it smaller than you think you should. Set a time and place. Stick to your promise like you would for someone you care about.</p><p>Because you do care about you! It just takes time to act like it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0a75a8f69e3e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-gravity/the-promises-we-keep-breaking-and-5-ways-to-start-keeping-them-0a75a8f69e3e">The Promises We Keep Breaking (And 5 Ways To Start Keeping Them!)</a> was originally published in <a href="https://medium.com/the-gravity">The Gravity</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[6 Chargeback Rules Every Online Seller Should Know Before The Holiday Rush]]></title>
            <link>https://blog.startupstash.com/6-chargeback-rules-every-online-seller-should-know-before-the-holiday-rush-67b2731352f1?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/67b2731352f1</guid>
            <category><![CDATA[money]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[fintech-startups]]></category>
            <category><![CDATA[banking]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Wed, 11 Nov 2026 20:31:01 GMT</pubDate>
            <atom:updated>2026-11-11T20:31:01.669Z</atom:updated>
            <content:encoded><![CDATA[<h4>The card networks tightened their limits in 2026. Here is what a dispute really costs, and how to keep your numbers safe.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iBnkCV_M8D51GtAE9Mi_Ww.jpeg" /><figcaption>Image is generated by AI</figcaption></figure><p>Picture this. It’s January, and you’re running a small beauty store online. You’ve had a successful couple of months (November and December). Suddenly the disputes start rolling in. One, then six, then nineteen in a single week.</p><p>Most of them come from customers who actually received their orders. One person even posted a picture of the product on their Instagram. Your bank then takes the money back, adds a fee, and asks you to prove you’re not lying.</p><p>This is known as a chargeback. If you sell anything online, you’ll run into one at some point. As this is the start of the holiday season, I wanted to share some of what I’ve picked up on the subject. The rules quietly changed in 2026 and many sellers may not have noticed.</p><p>Do you know your chargeback ratio from last month? If you don’t, then read on.</p><h3>1. Chargebacks Are Growing Faster Than Most Sellers Expect</h3><p>A chargeback happens when a cardholder asks their bank to reverse a payment. The bank then takes the money from the merchant, and the merchant must provide evidence to try to fight it.</p><p>These numbers are rising rapidly. A 2025 industry study published by Mastercard, based on Datos Insights research, put global chargebacks at 261 million for 2025. It expects 324 million by 2028, a jump of 24%. In money terms, that is $33.79 billion rising to $41.69 billion.</p><p>It isn’t just a forecast. In the same study, 59% of merchants said their chargeback volume had grown by more than 10% in a single year.</p><p>Why the rise? More card-not-present transactions, more subscriptions, and simply more people realising it’s easier to dispute a charge than to contact the store.</p><h3>2. Almost Half of Disputes Are Labelled as Fraud</h3><p>The most surprising thing is that merchants in that study said about 45% of their chargebacks were fraudulent. Banks say it’s even higher, at 72% of disputes.</p><h4>Two kinds of fraud</h4><p>Third-party fraud is when someone else has stolen a person’s card details. First-party fraud is when the person using the card actually made the purchase and then disputes it. This can be down to buyer’s remorse, a family member using the card, or simply abuse of the system.</p><p>Someone buys headphones from you, uses them for three weeks, then tells the bank they didn’t make the purchase. From the bank’s perspective this looks like card theft. From your perspective you’ve lost the product, the money and the fee.</p><h4>Why easy disputes mean more disputes</h4><p>The study also found that in the US, when banks allow customers to dispute through an app rather than a phone call, dispute volumes went up by 30% to 40%. Whilst this is great for genuine victims, it also opens the door for everyone else.</p><h3>3. Visa’s New Line in the Sand: 1.5%</h3><p>This is the change I most want sellers to know about.</p><p>In June 2025 Visa decided to combine its separate fraud and dispute monitoring programs into one, called the Visa Acquirer Monitoring Program, or VAMP.</p><p>The formula is simple. Take the number of fraud reports plus the number of disputes on your online card payments in a month. Divide that by the number of settled online card transactions.</p><p>When the program was first rolled out, a merchant was flagged as “Excessive” at a ratio of 2.2%. On 1 April 2026 that line was reduced to 1.5% in the US, Canada, Europe and Asia Pacific. However, the count rule still applies, so you are only flagged at that level if you also have at least 1,500 fraud reports and disputes in the month.</p><p>Let me make that real. Say your store processes 120,000 online Visa orders a month. At 1.5%, that is 1,800 combined fraud reports and disputes, so you would be above both thresholds. The fine is $8 for every one of them.</p><h4>“But I’m a small seller”</h4><p>Small merchants may never hit 1,500 events, but you’re not completely safe. If your payment provider exceeds its thresholds across its entire portfolio, fines of $4 or $8 per count can trickle down to merchants whose ratio is 0.5% or higher. So your ratio matters even when you are small.</p><p>One more thing to note: disputes settled through pre-dispute tools, and fraud claims that qualify under Visa’s newer evidence rules, are left out of the ratio. We’ll get onto that shortly.</p><h3>4. Mastercard Counts Differently, and the Fines Pile Up</h3><p>Mastercard also has its own program, and the maths is not the same.</p><p>A merchant enters the Excessive Chargeback Merchant level at 100 or more chargebacks in a month and a ratio of at least 1.5%. The higher tier starts at 300 chargebacks and 3%. To calculate your ratio, you divide this month’s chargebacks by last month’s sales.</p><p>That last point can cause issues. Say you had a busy December and a slow January. When you get to February you’ll still be getting chargebacks from December orders, but they are measured against January’s much smaller sales, so your ratio can spike even though nothing new went wrong.</p><p>The fines start small, at 1,000 dollars or euros in month two. If you don’t sort the problem out quickly, they climb to 100,000 a month at the first tier and 200,000 at the higher tier after 19 months. To get out, you need to be below the limits for three months in a row.</p><p>Can you see the pattern? Both networks reward the sellers who spot problems early, and punish the ones who wait.</p><h3>5. You Can Fight Back, If You Kept the Right Data</h3><p>In the past it has felt impossible to fight friendly fraud claims. Then, in April 2023, Visa introduced Compelling Evidence 3.0 for card-absent fraud disputes.</p><p>This makes sense. If a customer has bought from you before with no issues, using the same device or connection, it’s unlikely they’re a stranger.</p><p>To qualify, you need two earlier undisputed purchases from the same customer at your store, made between 120 and 365 days before the dispute. At least two details must match across all three orders: login ID, shipping address, IP address, or device ID. One of the two has to be the IP address or the device ID.</p><p>When Visa launched it in the UK, it estimated the rule could save retailers there about £31 million a year.</p><h4>The catch</h4><p>You can only use this if you stored the data. Many small stores don’t have device IDs or IP addresses saved with their orders, so when the dispute arrives the evidence simply isn’t there. This means the real work happens months before the chargeback, not after.</p><h3>6. The Clock Matters More Than You Think</h3><p>Most Visa disputes can be raised up to 120 days after the transaction is processed. For goods that never arrived, the window can start from the expected delivery date, up to 540 days.</p><p>This means a Black Friday sale on 27 November can come back to bite you in late March. Try to plan your cash, your stock of evidence and your team’s time accordingly.</p><p>And remember the hidden cost. Banks in the study spent between $9.08 and $10.32 just to process each dispute, and merchants wrote off 16% of their chargebacks as low-value cases not worth fighting. That money isn’t coming back.</p><h3>What to Check Before the Holiday Rush</h3><p>Here are the quick tips I would give any online seller this month:</p><ul><li>Find your Visa and Mastercard ratios for the last three months, and ask your payment provider how they calculate them.</li><li>Check that your billing name on card statements matches your store name. People will dispute what they don’t recognise.</li><li>Save IP address, device ID, login ID and shipping address with every order.</li><li>Send clear order and shipping emails with tracking.</li><li>Make refunds and cancellations easy. Refunds cost less than chargebacks.</li><li>Remind subscription customers before each renewal.</li><li>Ask about pre-dispute alerts. This means you can refund before a case becomes a chargeback.</li><li>Set aside time each week to review new disputes, not once a month.</li></ul><h3>The Takeaway: Treat Disputes Like a Number You Manage</h3><p>A chargeback is not just a lost sale. It is a signal, and in 2026 the card networks are watching that signal more closely than ever, with lower limits and real fines.</p><p>Going back to our skincare store, this time the owner checks her dispute numbers every Monday morning, right after her sales report. She stores device data for every order, has an easy refund button, and her billing name finally matches her brand. Her next January looks very different.</p><p>The good news is that most of this is in your control. Communicate clearly, keep good records and refund quickly.</p><p>So, before your biggest selling week of the year, would you have the evidence to hand if a flurry of disputes came through this January?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=67b2731352f1" width="1" height="1" alt=""><hr><p><a href="https://blog.startupstash.com/6-chargeback-rules-every-online-seller-should-know-before-the-holiday-rush-67b2731352f1">6 Chargeback Rules Every Online Seller Should Know Before The Holiday Rush</a> was originally published in <a href="https://blog.startupstash.com/">Startup Stash</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Goodbye, 16-Digit Card Number: 6 Things Merchants Should Know About Card Tokens Before 2030]]></title>
            <link>https://blog.startupstash.com/goodbye-16-digit-card-number-6-things-merchants-should-know-about-card-tokens-before-2030-3cd7937abda9?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/3cd7937abda9</guid>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[tokenization]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Mon, 09 Nov 2026 20:31:01 GMT</pubDate>
            <atom:updated>2026-11-09T20:31:01.585Z</atom:updated>
            <content:encoded><![CDATA[<h4>Why the card number is quietly leaving online checkout, and what that means for shops, apps and shoppers</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ndX3Dn3Ih9GvTiZoqT7S7w.jpeg" /><figcaption>Image generated with AI</figcaption></figure><p>Last month I was next to my aunt while she tried to buy a birthday gift on her phone. Out came the card from her handbag. Sixteen digits, typed with one thumb. Wrong expiry date, then a fix. Then the OTP. It came through, but by then the payment page had timed out.</p><p>Eventually she gave up and said she’d “do it from the computer later.” She never did.</p><p>As someone that works in payments, I see this happening all the time. But it made me think, as I’ve been reading for weeks about the removal of those long numbers on the front of your card from online payments. No big announcement, no sudden switch. It’s creeping in one checkout at a time, and hardly anyone seems to notice.</p><p>So I went digging. If you run a shop, an app or a subscription business, this one’s for you.</p><h3>1. The card number was never built for the internet</h3><p>The card number was originally created for a world of plastic, signatures and paper slips. Since then we’ve been entering these numbers into thousands of websites, keeping copies in shop systems and sending them across networks again and again.</p><p>Each copy is a risk. If the card number is kept in the shop’s system and they get hacked, your details are out there. Unlike a password, you can’t change your card number in seconds.</p><p>The issue is that we have been using one permanent secret as if it were safe to share.</p><p>And the cost is rising. Juniper Research expects online payment fraud losses to pass USD 91 billion by 2028. Yes, billion. I checked twice.</p><h3>2. A token is a stand-in that’s useless to thieves</h3><p>So what is replacing the card number? A token.</p><p>A token is essentially a random number that stands in for your actual card number. This means your card number isn’t sent to the shop when you pay. If a criminal were to steal a bunch of tokens in a data breach, they would be useless.</p><h4>One card, many tokens</h4><p>The clever part is that each device or merchant gets its own unique token for the same card. If you add your card to your phone and tablet, you’ll have a different token for each. If you save your card to a food delivery app and a travel website, each of those gets a different one too.</p><p>Let’s say you’re running a small online boutique selling sarees. You have stored cards to make things easier for your regular customers. With tokens, if someone were to hack in, the data cannot be used at the shop next door.</p><h4>A fresh code every time</h4><p>Every tokenized payment also carries a one-time code, a cryptogram. Copy it and try again? Nothing happens. It only works once.</p><h3>3. India already ran the experiment</h3><p>To see how this really works at scale, let’s look at India.</p><p>In January 2019 the Reserve Bank of India approved device tokenization, and in September 2021 it approved card-on-file tokenization. Then it took a huge step forward and said that from 1 October 2022, merchants, payment aggregators and payment gateways would no longer be able to store any card details.</p><p>There was some chaos around that deadline. Shops were panicking about renewals failing and customers were confused by the new consent pop-ups. But the system adapted.</p><p>By 31 December 2024 this had resulted in over 91 crore card-on-file tokens being created. They were used in over 320 crore transactions worth nearly Rs 11 lakh crore. And nearly 98% of e-commerce transactions were processed without the actual card data.</p><p>Ninety-eight per cent. That’s a huge number and it means this is definitely not a pilot anymore.</p><p>You will have noticed that Indian apps rarely ask you to enter your entire card number again. That is tokens at work.</p><h3>4. The 2030 deadline is real, and the world is halfway there</h3><p>India was driven by regulation, but globally the card networks are driving the change.</p><p>In June 2024 one of the two largest card networks declared its aim of 100% e-commerce tokenization in Europe by 2030. A few months later it went further and announced a plan to eliminate manual card entry and one-time or static passwords by 2030. That one-click goal has since been expanded to all markets. The other major card network says its aim is to tokenize 100% of e-commerce transactions.</p><h4>Where things stand</h4><p>The numbers are moving faster than I expected:</p><ul><li>By June 2025, nearly half of the first network’s online payments in Europe were tokenized. By June 2026 it was three in five.</li><li>Its one-click checkout is now available in 32 European markets, and returning customers make up over 70% of those checkouts.</li><li>The other network had more than 17.5 billion tokens by January 2026, over three times the number of physical cards.</li><li>On 28 October 2026, it said more than 60% of all e-commerce transactions on its network are tokenized, and that it processed tokenized payments in 200 countries and territories this year for the first time.</li></ul><h4>The death of guest checkout</h4><p>One of the most surprising stats I found was how far guest checkout, where you type your card details from scratch, has fallen: from 44% of that network’s online transactions in 2019 to about 16% in fiscal 2025. At its top 25 sellers it’s under 4%.</p><p>The big sellers have moved on. Will smaller businesses keep up?</p><h3>5. Tokens can mean more approved payments, not just less fraud</h3><p>Most people know tokens are a security measure, but that is only part of the story.</p><p>One network reports that tokenized online payments show, on average, 30% less fraud than payments made with the plain card number. It also sees a 4.6% increase in approval rates for card-not-present payments. Card-not-present just means the card isn’t physically there, so online or in-app payments.</p><p>Why would approval rates be higher? Well, the issuing bank has more confidence in a token than in a number that could have been stolen from anywhere. A token can also be updated when your card is replaced, so a saved payment keeps working after the old card expires.</p><p>If you run a meal-kit subscription business, you probably see renewals fail every month because customers have new cards and haven’t updated their details. Each failed renewal is a customer who might quietly leave. A few points more in approvals can mean many more people who keep getting their weekly box.</p><p>The same network says payment problems can cause up to 44% of abandoned digital purchases. One of those was my aunt!</p><h3>6. Passkeys are replacing the OTP, but tokens don’t stop every fraud</h3><h4>Goodbye to waiting for codes</h4><p>Even when the card number is protected, the customer still needs to prove they’re the real cardholder. For years that was done with an OTP, which is exactly what went wrong with my aunt’s purchase.</p><p>With payment passkeys you can use the fingerprint or face recognition on your phone. There’s no password to remember and no code to wait for. One network launched its passkey service in India, Singapore and the UAE first. In Europe, early one-click pilots ran faster and had three times fewer fraudulent transactions.</p><p>Say you’re squeezed into a crowded train. One glance at your phone and the ticket’s paid. No hunting for an SMS while the signal drops.</p><h4>Fraudsters are changing tactics</h4><p>To be completely honest, tokens and strong verification don’t end fraud. They move it.</p><p>Back in December 2025, the EU banking authority and the European Central Bank said payment fraud in the European Economic Area hit EUR 4.2 billion in 2024. The year before, it was EUR 3.5 billion. The fraud rate is still small, around 0.002% of value. For cards, strong customer authentication worked well. However, card fraud was 17 times higher when the payee was outside the EEA, where those checks aren’t required by law.</p><p>The report also flagged something worrying. A new kind of fraud is growing, where criminals trick people into approving payments themselves. No token helps when you’re the one tricked into pressing “approve.”</p><h3>What to check before 2030</h3><p>Running a shop, an app or a subscription service? Go through this list:</p><ul><li><strong>Raw card numbers still stored anywhere?</strong> Old databases, spreadsheets, support tickets, logs. Look in all of them.</li><li><strong>Network tokens from your payment partner?</strong> Ask. And ask if the cards you’ve already saved can be moved over.</li><li><strong>Approval rates.</strong> Are you tracking tokenized against plain card payments? The gap is worth seeing.</li><li><strong>Reissued cards.</strong> When a card changes, does the subscription keep renewing by itself?</li><li><strong>Guest checkout as default?</strong> Count the buyers who retype card details every time.</li><li><strong>Passkeys.</strong> If SMS codes are all you offer, find out what it takes to add more.</li><li><strong>Scams.</strong> Keep reminding customers not to share codes or approve payments they didn’t start.</li></ul><h3>The card number isn’t dead yet, but it’s packing its bags</h3><p>What struck me most is how quietly this is happening. Nobody has announced the end of the card number. It is gradually being replaced, one saved card at a time, until one day typing out 16 digits will feel as old as signing a paper slip.</p><p>Overall, tokens are a great way to keep your customers safe, can help more payments go through and reduce a big risk for your own systems. They’re no magic shield though, and the human side of fraud still needs watching.</p><p>Now I’m wondering: when did you last type your whole card number into a checkout? And if you run a business, how many of your customers still have to?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3cd7937abda9" width="1" height="1" alt=""><hr><p><a href="https://blog.startupstash.com/goodbye-16-digit-card-number-6-things-merchants-should-know-about-card-tokens-before-2030-3cd7937abda9">Goodbye, 16-Digit Card Number: 6 Things Merchants Should Know About Card Tokens Before 2030</a> was originally published in <a href="https://blog.startupstash.com/">Startup Stash</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How AI Is Transforming Fraud Detection and Why It Is Not a Perfect Solution]]></title>
            <link>https://medium.com/womenintechnology/how-ai-is-transforming-fraud-detection-and-why-it-is-not-a-perfect-solution-dff28d0b3d31?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/dff28d0b3d31</guid>
            <category><![CDATA[banking]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[fraud-detection]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[money]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Sat, 07 Nov 2026 15:21:01 GMT</pubDate>
            <atom:updated>2026-11-07T15:21:01.425Z</atom:updated>
            <content:encoded><![CDATA[<h4>From stolen passwords to deepfake voices: how banks are fighting back</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qpmXKpM0yeIM3WOXvoJPDA.png" /><figcaption>Image is generated by AI</figcaption></figure><p>Fraud is different from what it used to be. A year ago, it mainly involved something like a fake check or a stolen card. Now it can mean a person, a link that looks like it came from your bank or a robot trying thousands of passwords before you finish your coffee.</p><p>More people are using money, and bad people are following the money. The old ways of stopping fraud are finding it hard to keep up.</p><h3>Why old rule-based methods are not enough</h3><p>Banks have used rules for a time. If a transaction is above a certain amount, it gets flagged. If there are many failed attempts to log in or if a payment comes from a strange place.</p><p>Rules work. Only for a while. The problem is that bad people know how the rules work. If there is a limit of ₹50,000, they send ₹49,000. Break it into five smaller payments. They copy the habits of a customer. They use identities that seem real on paper.</p><p>Rules also cause problems for people. Anyone who has had their card blocked when they were travelling knows how it feels. You are standing at a counter in another country. The payment fails because the system saw something “strange.”</p><p>There is also the issue of numbers. Cards, UPI, wallets and online banking create millions of transactions every day. No one can check them all manually. A rule also can’t easily see the connections between a login from a device, a password change, and a group of strange transfers within ten minutes. Each one seems harmless on its own. Together they tell a story.</p><h3>What Artificial Intelligence does</h3><p>Instead of checking one thing at a time, an artificial intelligence model looks at many signs together: the amount, the time the type of business the device used the location, past actions and who the money is going to. It learns what normal looks like for each person and notices when something is different.</p><p>Imagine a customer who usually spends in their city. One night, their account logs in from a phone, and soon after, a large amount goes to someone they have never paid. A rule might only see the amount. An AI model sees the situation and marks it as high risk.</p><p>What happens next depends on the score. A low-risk payment goes through. A strange one might ask for an OTP or a quick check. A high-risk one gets stopped for a human to check. All of this happens in seconds, which’s important for instant payments because once the money is gone it is hard to get back.</p><p>AI models can also find fraud that no one has seen before. Anomaly detection does not need a known pattern. It just shows when something does not fit, like a group of transactions from many accounts using one device or large transfers right after a password change.</p><p>An anomaly is not proof of fraud, though. Real people do things all the time. The best systems treat these flags as a reason to look closely, not as a final decision.</p><h3>Where it helps</h3><p>Card and payment fraud. AI models can spot when someone is using a card without permission by comparing each transaction with the customer’s habits and the merchant’s reputation. The tricky part is balance. Block too much, and customers get upset. Block little and losses add up.</p><p>Fake identities. Fraudsters often mix fake details to create a person who does not exist. AI can find differences across applications, devices and account activity. Biometrics and checking documents add another layer. They shouldn’t be the only one.</p><p>Account takeover. A stolen password alone should not be enough. If the login pattern, device and behaviour all seem wrong, the system can ask for checks or limit what the account can do.</p><p>Money laundering. Laundering rarely appears in one transaction. It appears in networks: accounts passing money through middlemen before it ends up somewhere common. Graph analysis can find these links. Still, an alert is not proof of a crime. Banks must still. Report properly under rules against money laundering.</p><p>Tricks. Many losses start with someone being tricked, not a system being hacked. Technology can find messages and sudden changes in behaviour, but customer awareness is just as important. No algorithm stops someone who’s determined to follow a scammer’s directions.</p><h3>Generative AI has both good and bad sides</h3><p>Fraud teams can now use AI assistants to summarise alerts and organise investigations. Explain why a payment was flagged. That helps experienced analysts focus on the cases.</p><p>Bad people have the tools. Phishing emails used to be easy to spot because of spelling mistakes, and those days are gone. Fake documents, copied voices, and deepfake videos are getting more real. A phone call that sounds like your manager is no longer proof of anything.</p><p>So banks can’t rely on one way to protect. A better plan uses more than one:</p><p>independent checking of identity</p><p>extra checks for important actions</p><p>watching behavior</p><p>limits on payments and checking risks</p><p>training for staff and customers</p><p>human checks for high-risk cases</p><p>AI is one part of this plan, not the whole plan.</p><h3>The problem of alerts</h3><p>Every team that deals with fraud has this challenge. A real payment that gets blocked makes the customer upset and adds work for support teams. Can cost a merchant a sale. If the rules are too loose, fraud losses go up.</p><p>AI helps because it looks at the situation instead of just one limit. A large payment from a long-time customer on a trusted device is seen differently from the same amount on a new device.</p><p>Fewer alerts don’t automatically mean a better model. It might just be letting fraud in. Teams should track losses, how well they find fraud, how false alerts they get how much customers are upset how much work analysts have and how quickly things get fixed together not just one at a time.</p><h3>The risks that no one should ignore</h3><p>Quality of data and fairness. A model is only as good as the data it learns from. If the old data is missing or not balanced, the model may unfairly see some customers or areas as risky. This needs checking</p><p>Understanding why. When a system freezes someone’s account, the bank must explain why. Saying “the model said so” is not enough, especially when people need that money.</p><p>Bad people trying to trick the system. Fraudsters test systems, learn their limits. Change their ways. Models also lose effectiveness as tactics change, so they need checking and updating.</p><p>Privacy. Good detection means looking at financial and personal data. That needs rules about who can see it, how it is kept safe, how much is collected and how long it is kept. The goal is to protect customers, not to watch them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cqBdEFXvFZWOR610b-2ZSA.png" /><figcaption>Image is generated by AI</figcaption></figure><h3>How to use it in life</h3><p>Buying a smart model is not a plan. In my view, the right way is like this:</p><p>First, find your risks. Is it card fraud, account takeover, fake identities or fake accounts? Clear goals shape everything</p><p>Fix your data. Messy or not consistent records will stop the best model.</p><p>Use what you already have. AI should work with the rules, case tools and checks, not replace them all at once.</p><p>Test before using it. Check how well it finds fraud, how wrong alerts are, how fast it is and how it acts when data is missing or someone is trying to trick it.</p><p>Keep people in the loop. Hard cases and decisions still need workers. Keep changing because fraud is always changing.</p><h3>Where things are going</h3><p>Some trends are clear. Decisions will be faster and more detailed with systems looking at signs at once. Banks, payment companies and sellers will probably share fraud information, which needs good legal and data rules to work safely. As fake videos get better, checking will use more than one way to prove identity, not just one voice, photo or document, especially for big transfers and changes to account details.</p><p>Regulators and customers will also want AI systems to be clear, checked and fair. Banks that build these habits early will do better than those who add them later.</p><h3>The Main Insight</h3><p>AI is making fraud detection faster and more accurate. It finds patterns people miss and reacts in time. It also brings its own issues: bias, privacy, being able to explain and bad people who change their methods.</p><p>The organisations that do well will not be the ones with the advanced model. They will be the ones that use AI with data, multiple layers of security, smart people and constant checking.</p><p>In the end, this is not about stopping losses. It is about trust. People will use money only if they believe their money is safe. Technology can help build that trust. How banks use it will decide if they keep it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dff28d0b3d31" width="1" height="1" alt=""><hr><p><a href="https://medium.com/womenintechnology/how-ai-is-transforming-fraud-detection-and-why-it-is-not-a-perfect-solution-dff28d0b3d31">How AI Is Transforming Fraud Detection and Why It Is Not a Perfect Solution</a> was originally published in <a href="https://medium.com/womenintechnology">Women in Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[One Account, Hundreds of IBANs: 6 Lessons on Virtual IBANs for Payment and Fintech Firms]]></title>
            <link>https://medium.com/@pathakindukumari/one-account-hundreds-of-ibans-6-lessons-on-virtual-ibans-for-payment-and-fintech-firms-fbf1fb1cea2d?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/fbf1fb1cea2d</guid>
            <category><![CDATA[iban]]></category>
            <category><![CDATA[banking]]></category>
            <category><![CDATA[money]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[fintech]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Sat, 07 Nov 2026 07:01:32 GMT</pubDate>
            <atom:updated>2026-11-07T07:01:32.145Z</atom:updated>
            <content:encoded><![CDATA[<h3>Virtual IBANs make it much easier to know who paid you. But from 2027, the EU wants to know who’s using every single one of them, and that changes things.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r2C79JWY1loGEKu0SNMhfQ.png" /></figure><p>Picture a small online marketplace in Lisbon. It has 400 sellers, a single bank account and a finance person called Sofia.</p><p>Every morning Sofia opens the bank statement and finds dozens of incoming transfers. Some say “invoice 2231”. Some say “payment”. One just says “for the blue sofa”. She spends half her day working out who paid what.</p><p>Now give that same marketplace virtual IBANs. Every seller and every buyer gets their own account number. Money arrives, and the system already knows who sent it and what it’s for. Sofia gets her mornings back.</p><p>That is the promise of virtual IBANs. I’ve been reading a lot about them recently, including the European Banking Authority’s report and the new EU anti-money laundering rules, and honestly they’re a bit more complicated than the sales pitch. So here are 6 lessons I’ve picked up.</p><h3>1. What a virtual IBAN actually is</h3><p>An IBAN is the long account number you use to send money across Europe. It usually points to one real bank account.</p><p>Now a virtual IBAN, or vIBAN if you’re in a hurry, is basically a lookalike. Same format, same length, same country code at the front. The difference is that there’s no separate account behind it. Payments sent to it are routed into a “master” account that holds the actual money.</p><p>So one business can have one real account and hundreds, sometimes thousands, of virtual IBANs sitting on top of it. Each one can be given to a customer, a supplier, a subsidiary or a single invoice.</p><p>The EBA made a point I found really interesting. Virtual IBANs have the same functionality and format as normal IBANs, so the person sending money usually can’t tell the difference. They type it in, hit send, done.</p><h3>2. Why finance teams keep asking for them</h3><p>Reconciliation. That’s most of it, really.</p><p>Month-end in a lot of accounts teams still means a spreadsheet, a cold coffee and a list of payments where the customer forgot the reference (or typed their dog’s name in it). Give each payer their own number and most of that work just goes away. You don’t have to rely on someone typing the right reference, because the account number itself tells you who paid.</p><p>There are some other reasons too:</p><ul><li><strong>Local presence without local branches.</strong> A company can offer customers a local-looking IBAN in a country where it doesn’t have a bank account.</li><li><strong>Fewer real accounts to manage.</strong> Treasury teams can run a big group from one or two master accounts instead of dozens.</li><li><strong>Cleaner client money.</strong> Payment firms can give each client their own number while still keeping funds together in a safeguarding account.</li></ul><p>Big banks have been doing this for corporate clients for years. J.P. Morgan, Citi and HSBC all sell virtual account products. Thailand’s Krungthai Bank launched one for its business customers in February 2026. What’s changed is that fintechs and payment firms now hand out virtual IBANs to smaller businesses and even individuals.</p><h3>3. The EBA’s warning: nobody can see who is behind the number</h3><p>In May 2024 the EBA published its first report on virtual IBANs. It came out of a fact-finding exercise with national supervisors across the EU, and it’s worth a read if you work anywhere near this area.</p><p>The main concern is simple. If the person paying can’t tell a virtual IBAN from a normal one, and the bank holding the master account only sees the business that owns it, then who actually knows the end user?</p><p>The report listed a range of risks. A few stood out to me:</p><ul><li><strong>Unknown end users.</strong> The people using a vIBAN may not be known to every payment provider in the chain.</li><li><strong>Different rules in different countries.</strong> A vIBAN can carry one country’s code while the master account sits in another. So which country’s supervisor is actually in charge? Often each one reads the AML rules its own way, and things fall through the cracks.</li><li><strong>Confusing for consumers.</strong> A customer may think their money is protected by a bank in one country, when it’s actually held somewhere else, sometimes not even in a bank.</li><li><strong>Regulatory arbitrage.</strong> Firms could pick the country with the lightest supervision.</li></ul><p>None of this means virtual IBANs are bad. It means the convenience comes with responsibility.</p><h3>4. The new EU rules arriving in 2027 and 2029</h3><p>This is where things get serious.</p><p>The EU’s new Anti-Money Laundering Regulation (AMLR) is the first EU law to deal with virtual IBANs directly. Most of it applies from 10 July 2027. Under it, the bank or payment firm that services the master account must be able to get information on the people using the virtual IBANs, and fairly quickly if asked.</p><p>The companion directive (often called AMLD6) goes further. Countries must add virtual IBANs to their central bank account registers. Each entry will include the virtual IBAN, the account it routes to and the opening and closing dates. National registers then have to be linked up through a new EU system called BARIS by July 2029.</p><p>In plain words: a virtual IBAN will no longer be invisible to the authorities. If a payment goes through one, investigators will be able to trace where it really landed.</p><p>Germany’s regulator BaFin didn’t wait. It published a supervisory statement on virtual IBAN risks in 2026 setting out what it expects from firms until the AMLR applies. I’d expect other supervisors to follow.</p><h3>5. Name checks don’t always see the real customer</h3><p>In my last story I wrote about Verification of Payee (VoP), the name check that’s now part of every euro payment. You type a name and an IBAN, and the bank tells you if they match.</p><p>Virtual IBANs make this tricky.</p><p>Say you pay a supplier using a virtual IBAN given to them by their payment provider. The name check might return the name of the payment provider, or the marketplace, rather than the supplier you expected. You get a close match or a no match, even though everything is genuine.</p><p>The Euro Banking Association flagged this in 2025. It pointed out that a VoP check may only cover the main account and not the sub-accounts sitting underneath it.</p><p>This matters a lot, because a confusing name check trains people to ignore warnings. And once people ignore warnings, the whole point of VoP starts to disappear.</p><p>If you issue virtual IBANs, think carefully about what name comes back when someone checks it. Ideally it should be the name your customer’s payers expect to see.</p><h3>6. What to check before you issue your next virtual IBAN</h3><p>If you run a fintech, a marketplace or a payment business, here’s my short list.</p><p><strong>Before you issue</strong></p><ul><li>Know your end customer, not just the business that owns the master account.</li><li>Record legal names and trading names properly. Bad data shows up later as failed name checks.</li><li>Be clear with customers about where their money is actually held and how it’s protected.</li></ul><p><strong>While it’s in use</strong></p><ul><li>Monitor payments per virtual IBAN, not just at master-account level. A mule account can look perfectly boring once it’s mixed in with 900 others.</li><li>Write down who owns each number, the day it went live and the day you shut it. Sounds dull, I know. You’ll need it for the registers anyway.</li><li>Test what Verification of Payee returns for your numbers. Try it yourself.</li></ul><p><strong>Before July 2027 sneaks up on you</strong></p><ul><li>List every product that hands out vIBANs, and which country codes they use. You might be surprised.</li><li>Ask your banking partner a blunt question: if a regulator wants the user behind vIBAN X by Friday, can you get it?</li><li>Find out who’s filing what with the central registers. Is it them, you, or nobody yet?</li></ul><h3>Who’s behind the number, then?</h3><p>Virtual IBANs solve a real, everyday problem. Anyone who has spent hours matching payments to invoices will understand why businesses love them.</p><p>But a number that looks like a bank account and behaves like one should come with the same transparency. That’s exactly where regulators are heading. Give it until 2029 and every vIBAN in the EU should sit in a register, with a name attached.</p><p>When I first picked up the EBA report I half expected forty pages of back-office plumbing. Instead it kept circling back to onboarding, fraud and trust, the stuff customers actually feel.</p><p>So here’s where I’ve landed. Sort it out this year while it’s still a product decision. Leave it, and it turns into a compliance fire drill in summer 2027, probably during somebody’s holiday. Names, real customer checks, boring tidy records. That’s honestly most of the job. Sofia would agree, I think, and your compliance lead definitely will.</p><p>Over to you. If you’ve rolled out vIBANs, did reconciliation actually get easier, or did the mess just move to another desk?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fbf1fb1cea2d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Does The Name Match?]]></title>
            <link>https://blog.startupstash.com/does-the-name-match-c7aba0f4fb98?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/c7aba0f4fb98</guid>
            <category><![CDATA[banking]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[fraud-prevention]]></category>
            <category><![CDATA[payments]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Fri, 06 Nov 2026 17:26:01 GMT</pubDate>
            <atom:updated>2026-11-06T17:26:01.573Z</atom:updated>
            <content:encoded><![CDATA[<h3>Does the Name Match? 6 Lessons From Europe And The UK on Stopping Payment Scams Before The Money Moves</h3><h4>Instant payments now check the payee’s name, and banks are paying back scam victims. Here is what the 2025 and 2026 rules teach payment and fintech firms.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*379_iMroZePImVvpgCm1tg.png" /></figure><p>Imagine you’re a small business owner and are about to pay a supplier. You receive an email saying they’ve changed their bank details. The invoice looks genuine, and the amount you need to pay is correct, and you need to pay quickly.</p><p>You update your account number and hit ‘send’. The payment has left your account and been received by a fraudster in seconds.</p><p>Now imagine you’re in exactly the same situation, BUT your bank has performed a name check for the recipient. So now you’ve got a reason to pause and think.</p><p>This is one of the largest changes to payments in Europe. Regulators across Europe and the UK are looking to move away from leaving customers responsible for identifying scams and are placing more responsibility, and sometimes financial responsibility, onto payment providers. Here are 6 lessons I’ve learnt from all this.</p><h3>1. Why instant money needed a name check</h3><p>Payments became faster but caused issues.</p><p>Under the EU Instant Payments Regulation (EU) 2024/886, banks and payment firms in the euro area had until 9 January 2025 to start receiving SEPA instant payments. Sending them, with Verification of Payee switched on, came next, by 9 October 2025. And the money has to land within 10 seconds, for no more than a normal transfer costs.</p><p>How can we make payments faster without increasing fraud?</p><p>One way is to ensure that you’re paying the correct person. From 9th October 2025 Verification of Payee (VoP) has applied the same way throughout Europe. This performs a name check of the payee and ensures their name matches the account number (IBAN).</p><p>If the name does not match, our business owner will notice an alert. It’s not a guarantee. But the point is to stop it before the money moves, not after someone reports it missing.</p><h3>2. What Europe’s payee checks show after one year</h3><p>In August 2026, data was published by the European Payments Council showing that there were around 2.2bn verification requests during Q2 2026. At this rate there would be over 8.8bn checks per annum. 95% were answered in less than a second.</p><h3>Three results, three different decisions</h3><p>The results from Q2 2026 were:</p><ul><li>Match: 66.2%</li><li>Close match: 16.2%</li><li>No match: 15.4%</li><li>Not applicable: 1.4%</li><li>Error: 0.8%</li></ul><p>A match means the name and IBAN line up. A close match means they’re nearly the same, so it’s worth a second look. No match means the name doesn’t fit the account at all.</p><p>Something to note is the number of ‘close matches’. This can happen if the legal name of your supplier does not exactly match the trading name on invoices. This should be looked at carefully, not just rejected or waved through.</p><p>The ‘no match’ result should be treated with extra care. This may be due to a typo or someone has entered their own details.</p><h3>People are noticing</h3><p>During the period between October 2025 and end of February 2026 there were 57.4 million verification requests processed through Latvia’s service. During February 2026 a survey was carried out and found that 68% were aware of the name check. Half of those stated that it had helped to avoid incorrect transfers.</p><p>This is my favourite statistic!</p><h3>What comes next</h3><p>Version 2.0 of the VoP rulebook will be published in November 2026 and go live in November 2027 (subject to the EU’s new Payment Services Regulation). There are proposals to extend VoP to cover non-euro schemes within the European Economic Area.</p><h3>3. The UK made banks pay back scam victims, and fraud fell</h3><p>The UK went one step further by requiring payment service providers to cover losses incurred by customers falling victim to authorised push payment (APP) scams.</p><p>An authorised push payment scam tricks you into sending your funds. For example, a grandmother gets a call from ‘her bank’ stating there is a threat to her account and that she needs to transfer her savings to a secure account. She transfers the money.</p><h3>What changed on 7 October 2024</h3><ul><li>Sending payment providers have to refund eligible victims on Faster Payments and CHAPS, generally within 5 business days.</li><li>Maximum refund of £85,000 per claim. Optional excess up to £100.</li><li>The receiving payment provider pays 50% of the cost.</li><li>Claims can be refused if made more than 13 months after the payment.</li></ul><p>This creates an incentive for the recipient bank to identify mule accounts.</p><h3>What the results tell us</h3><p>Between 7 October 2024 and 31 March 2026, 88% of funds lost to in-scope APP scams (around £316 million) were refunded according to the Payment Systems Regulator dashboard. In 2024 UK Finance reported a 61% refund rate however this isn’t directly comparable. Around 438,300 claims were made, of which around 301,500 were in-scope. 82% were resolved within 5 business days.</p><p>Behaviour change? An independent review carried out in 2026 by Frontier Economics estimated losses from APP fraud had reduced by £73 million per annum. There was also a reduction in almost 35,000 scams. Refunds across all claims went up from 54% to 65%, and 97% for in-scope claims.</p><p>When fraud is affecting both parties of a payment, preventing fraud becomes a business case rather than an issue for the support desk.</p><h3>4. When scammers pretend to be your bank</h3><p>A name check can help when dealing with a fraudulent account but what if they are pretending to be the bank?</p><p>You receive an email with your bank branding, telling you there is a problem with your security or you get a call from a number which appears familiar, asking you to urgently make a payment to save your account. You believe this and allow the payment to go ahead.</p><p>Although it is not live yet, the EU reached a provisional deal on the Payment Services Regulation (PSR) and PSD3 on 27 November 2025. This means:</p><ul><li>Providers must check the payee name matches the identifier, and if not, refuse the payment and inform the payer.</li><li>Any payment started or changed by a fraudster counts as unauthorised.</li><li>If the scammer has impersonated the consumer’s own provider and the consumer was scammed into making a payment this must be fully refunded by the provider, provided the consumer reports it to both the police and the provider.</li><li>Online platforms can be liable to providers if they don’t remove fraudulent content that has been reported.</li></ul><p>Article 59 states that the provider must refund or provide a reasoned refusal within 15 business days. It only applies to consumers and only when their own provider has been impersonated using spoofed phone numbers, emails, websites or Apps. And it’s the provider who has to prove the customer was grossly negligent.</p><p>For fintechs, fraud prevention now includes helping customers tell your real messages from fake ones.</p><h3>5. What changes for businesses and fintechs</h3><p>The biggest mistake would be to think that all your customers will be protected. If your employee gets scammed by someone pretending to be your bank, don’t expect the same EU refund right. It’s for consumers, not companies.</p><h3>Business customers need their own checks</h3><p>Imagine a member of staff in finance receives an email from a supplier saying they need to update their bank details. During a busy month end, VoP could flag a mismatch, but there is no way of knowing if the person requesting the change is genuine.</p><p>Call the supplier on a number you have on file, rather than the number in the email. Having a matching name is useful evidence but does not guarantee it is genuine.</p><h3>Close matches start with good onboarding</h3><p>If a business has different legal and trading names and you just have one name recorded then you could get some confusion. Payment companies should ensure they have:</p><ul><li>Legal names</li><li>Trading names</li><li>Account-holder names for individuals</li><li>The same records across all their systems</li></ul><h3>6. What to check before your next payout run</h3><h3>Before accepting instructions</h3><ul><li>Get the beneficiary names and account details right the first time.</li><li>Record their legal and trading names.</li><li>Say what match, close match and no match mean in plain words.</li></ul><h3>Before sending payouts</h3><ul><li>Have a plan for a no match. Who looks at it, and what then?</li><li>Show customers enough to judge a near-match themselves.</li><li>Never assume that if a name passes then it is genuine.</li><li>Always independently verify any changes to a supplier’s bank details.</li></ul><h3>After a suspected scam</h3><ul><li>Make it easy for people to tell you something went wrong.</li><li>Hang on to the payment records and the messages around it.</li><li>Know which refund rules apply to you, and how long you’ve got.</li></ul><p>One last thing for the bosses. Are you counting the dodgy payments you caught, or the customers you kept safe before the money left?</p><h3>So, does the name match?</h3><p>Europe is starting to include name checks within payments. The UK has shown that forcing refunds will encourage payment providers to stop scams. The EU also now has new rules surrounding impersonation.</p><p>However there is more to do. You still need clean customer data, logical checks and a calm approach when something seems odd. Customers should feel like they are being protected by the warning.</p><p>The main thing I took away was payments should be secure from the start. It is brilliant that payments can be made into an account in seconds however, it needs to be the correct person.</p><p>Do you think payment companies should improve name checks, detect bank impersonation or be held accountable for scams?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c7aba0f4fb98" width="1" height="1" alt=""><hr><p><a href="https://blog.startupstash.com/does-the-name-match-c7aba0f4fb98">Does The Name Match?</a> was originally published in <a href="https://blog.startupstash.com/">Startup Stash</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DORA’s First Report Card: 6 Lessons For Payment And Fintech Firms From 3,383 Major IT Incidents]]></title>
            <link>https://blog.startupstash.com/doras-first-report-card-6-lessons-for-payment-and-fintech-firms-from-3-383-major-it-incidents-f3ac3c25bf9f?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/f3ac3c25bf9f</guid>
            <category><![CDATA[regulatory-compliance]]></category>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Thu, 05 Nov 2026 20:21:01 GMT</pubDate>
            <atom:updated>2026-11-05T20:21:01.494Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n4JFuMWajeQTgxWnv_hd-A.jpeg" /><figcaption>Image generated with AI</figcaption></figure><p>Picture yourself at a supermarket checkout in Madrid. Your basket is full, your banking app works and there is plenty of money in your account. Then your card payment fails. Your bank has not been breached. The payment terminal has just lost its connection.</p><p>Now imagine that happening to thousands of people at once.</p><p>That nearly happened in April 2025, when the blackout hit Spain and Portugal. At the time I was deep in the EU’s Digital Operational Resilience Act (DORA), and one idea kept nagging me. <strong>A bank or payment firm can be badly disrupted without any cyberattack at all.</strong></p><p>The first EU-wide report on major IT incidents under DORA reflects this. Cybersecurity incidents made up just 10% of the major ones in 2025.</p><p>Work in payments, e-money, crypto or fintech? Then these findings matter to you. Almost two years into DORA, we can finally watch it meet real life. Here are six lessons that stood out to me.</p><h3>1. DORA Is Not Just Another Regulation for Banks</h3><p>DORA (Regulation (EU) 2022/2554, if you want the formal name) went into the EU Official Journal on 27 December 2022. The date that counts is 17 January 2025, when it started to apply.</p><p>I was surprised by how broad it is. It covers 20 types of financial entities and the ICT (information and communication technology) third-party service providers they rely on.</p><h4>Who needs to pay attention?</h4><p>In scope are payment institutions, electronic money institutions, account information service providers, crypto-asset service providers and issuers of asset-referenced tokens. Banks, investment firms and insurers too, among others.</p><p>Take a small payment startup. A lean team, a few key functions run by outside providers. When that technology stops, customers simply cannot pay or get paid. DORA expects your firm to manage that risk, not the provider.</p><p>Five areas sit under the regulation. These are ICT risk management, incident management and reporting, digital operational resilience testing, ICT third-party risk (with oversight of critical providers) and information sharing.</p><p>My main takeaway? Forget paperwork for the inspector. DORA is about keeping the doors open, and getting back up fast when technology breaks, even when someone else runs part of it.</p><h3>2. The Four-Hour Incident Reporting Clock Changes Everything</h3><p>A firm cannot wait for an IT problem to be fixed before it tells its supervisor. For a major ICT incident the deadlines are tight:</p><ul><li><strong>Initial notification:</strong> due within 4 hours once the incident is classified as major, or within 24 hours of detection, whichever is earlier.</li><li><strong>Intermediate report:</strong> within 72 hours of the initial notification.</li><li><strong>Final report:</strong> within one month.</li></ul><p>Firms must also report new contracts with ICT providers at least once a year. Significant cyber threats can be reported voluntarily.</p><h4>Why does this matter for a small fintech?</h4><p>Imagine an e-money firm hit by a system failure one morning. Customers cannot log in and payments are failing. The tech team is still hunting for the cause, and the bosses want answers now.</p><p>Who decides if this incident counts as major? Who gathers the facts, and who contacts the supervisor before the clock runs out?</p><p>Nobody should be answering those questions for the first time in the middle of an outage. Incident management is a business job, not only an IT job. Technical detail can wait, since the later reports exist for that. Ownership cannot. Without it, even a strong team can blow the deadline.</p><h3>3. The First DORA Report Shows Most Problems Were Not Hackers</h3><p>The EBA, EIOPA and ESMA, the three European Supervisory Authorities, released the first EU-wide report on major DORA incidents on 3 June 2026.</p><p>Financial entities reported <strong>3,383 major ICT incidents</strong> in 2025. That works out at about 282 a month, or 0.18 per firm. More than 60% came from the banking sector and 16% from payments. Around a third (1,056) had a cross-border impact, and in about 8% of all major incidents more than 10 countries were affected.</p><h4>System failures, not cyberattacks, led the list</h4><p>System failures accounted for 51% of incidents, external events for 27% and payment-related incidents for 18%. Only 10% were cybersecurity incidents. Of those, 33% were DDoS attacks (floods of traffic that knock a service offline) and 31% involved data theft or manipulation, including identity theft.</p><p>That 10% figure really caught my eye. A software fault can hurt daily operations as much as an attacker can.</p><p>The root causes tell a similar story. Half were system failures or malfunctions. Close to a third, 29%, came from a failure at a third-party provider, and human error caused 12%.</p><p>The authorities warned firms to watch their third-party dependencies, “even those not designated as critical.” They added that cybersecurity must get better as AI-driven tools spread. My reading? Give a function to someone else and you still carry the fallout when it fails.</p><h4>Small numbers can hide reporting gaps</h4><p>In almost 60% of major incidents, fewer than 1,000 clients were affected, or none at all. Two thirds touched either no transactions (32%) or fewer than 1,000 (26%). Only 30 incidents, about 1%, hit more than a million transactions, mostly in banking and payments.</p><p>This is the most telling part. Half of the incidents reported no cost at all, or a negligible one below EUR 1,000. The authorities said this may point to incorrect reporting. A low loss figure does not make a small incident.</p><h3>4. Two Real Outages in 2025 Show What Resilience Means</h3><h4>February: a settlement system goes dark</h4><p>On 27 February 2025, the Eurosystem’s TARGET services suffered a major incident caused by a very rare storage hardware fault. T2 was unavailable for about 10 hours and T2S for about 8 hours. TIPS instant payments were partly disrupted for about an hour. Some salary and pension payments were delayed.</p><p>Someone waiting for their salary does not care about storage hardware. They just expect the money to arrive.</p><h4>April: a blackout stops everyday payments</h4><p>On 28 April 2025, a total blackout hit Spain and Portugal for about 10 hours. The banks’ data centres kept running on backup power. But the internet was down, so apps and websites could not be reached, and card terminals in shops lost power or connectivity. Ordinary payments stopped.</p><p>This breaks a common belief that if the data centre is up, the service is up. Think of what a payment needs: bank, network, terminal, power. One link snaps and the whole chain goes.</p><p>So what is your plan if your systems are running but customers cannot reach them?</p><h3>5. Critical Technology Providers Are Now Under Direct EU Oversight</h3><p>The ESAs named 19 critical ICT third-party providers on 18 November 2025. Those providers, mostly in cloud, data, telecoms and IT services, now sit under direct EU oversight. The list gets reviewed every year.</p><p>How were they picked? The authorities used the registers of information that financial firms submit about their ICT contracts.</p><p>There is a practical lesson hiding here. Third-party risk is unmanageable if you do not know who you depend on, or what for. A crypto firm with an incomplete register has a gap in its risk management and its reporting.</p><p>Still, a provider can cause real harm without any critical label. Third-party failures sat behind 29% of the major incidents in 2025.</p><h3>6. What Supervisors Are Still Finding Weak</h3><p>In August 2026, the Dutch markets regulator, the AFM, shared where firms are actually going wrong.</p><p>Registers of information were a struggle at first. The EBA accepted only 40% in 2025. By 2026 that figure had climbed to 94%. It is real progress, but it shows how hard accurate contract data is.</p><p>The AFM received 123 DORA notifications in 2025: 65 major incidents, 54 new ICT contracts and 4 significant cyber threats. It said incident numbers were lower than expected and urged firms to review their incident processes. A low count may simply mean problems are not being spotted.</p><p>Other findings were just as practical:</p><ul><li>Many firms lacked required policies. Group-level policies are allowed, but the licensed entity stays fully responsible.</li><li>Detection was generally good, but prevention was weak, especially logical access management and patch and vulnerability management.</li><li>The first threat-led penetration tests (TLPT), which copy the methods of real attackers, began in 2025. Under DORA Article 26, certain firms must run them at least every 3 years.</li></ul><p>A great response team cannot make up for access rights nobody reviews or weaknesses nobody patches.</p><h3>What to Check Before Your Next Outage</h3><p>Based on the rules, the first EU-wide report and the AFM’s findings, here is a short checklist I would run through:</p><ul><li><strong>Reporting owner:</strong> Does everyone know who classifies an incident as major and who files the first notification within 4 hours of that call (or 24 hours of detection)?</li><li><strong>Contract register:</strong> Is every ICT provider and contract in one accurate, current register?</li><li><strong>Dependencies:</strong> What happens if your main technology or payment partner goes down?</li><li><strong>Access and patching:</strong> Are access rights reviewed and known vulnerabilities fixed on a set schedule?</li><li><strong>Policies:</strong> Do they reflect how your licensed entity really operates?</li><li><strong>Recovery tests:</strong> Have you tested recovery lately, in a messy, realistic scenario?</li></ul><p>Ask them on an ordinary Tuesday. Not the week before an inspection.</p><h3>The Biggest Lesson: Resilience Is About More Than Technology</h3><p>After reading DORA and the first EU-wide incident report, one idea stays with me. <strong>Attackers are not always the biggest risk in financial services. System failures, outside events, human mistakes and the everyday providers firms rely on often do more damage.</strong></p><p>DORA has made those responsibilities visible. But policies and registers are only the start. A payment firm needs to know what happens when transactions stop. An e-money firm needs to know how customers will reach their money. And the vendors behind them need to see how their failures reach ordinary people.</p><p>The February settlement outage and the April blackout proved it. Nobody attacked anyone, yet ordinary people felt the pain.</p><p>Resilience is not measured by how many policies you have written. It is measured by what happens when things go wrong.</p><p><strong>So, if your most important technology service went down tomorrow morning, would your business know exactly what to do next?</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f3ac3c25bf9f" width="1" height="1" alt=""><hr><p><a href="https://blog.startupstash.com/doras-first-report-card-6-lessons-for-payment-and-fintech-firms-from-3-383-major-it-incidents-f3ac3c25bf9f">DORA’s First Report Card: 6 Lessons For Payment And Fintech Firms From 3,383 Major IT Incidents</a> was originally published in <a href="https://blog.startupstash.com/">Startup Stash</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Where Is Your Customers’ Money Right Now? 6 Safeguarding Lessons for Payment and Fintech Firms]]></title>
            <link>https://medium.com/@pathakindukumari/where-is-your-customers-money-right-now-6-safeguarding-lessons-for-payment-and-fintech-firms-723d9bede052?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/723d9bede052</guid>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[cash]]></category>
            <category><![CDATA[banking]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[money]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 11:49:49 GMT</pubDate>
            <atom:updated>2026-11-03T11:49:49.800Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>A balance on a screen is not the same as money protected in a bank. Here is what the UK, EU, US and India are teaching us about safeguarding customer funds in 2025 and 2026.</em></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JtS025fxzqQg5mJ4iQsiCA.png" /><figcaption>Image is generated by AI</figcaption></figure><p>Imagine you’re a small online trader looking to pay for your next batch of stock by viewing your payment app on a Monday morning. The dashboard shows you have a nice balance.</p><p>You click withdraw and nothing happens. The platform has gone down and the partner bank is trying to determine whose money is whose. You can see you have a balance but can’t access it.</p><p>Where is your money?</p><p>This should be easy to answer if you’re running a payment, e-money or embedded finance company. For me, safeguarding is being confident that you know where your customers’ money is, can prove ownership and can retrieve the money in the event of anything going wrong. Due to the changes in rules in 2025 and 2026, here are 6 lessons which I think all fintech professionals should know.</p><h3>1. A Balance on a Screen Is Not Money in a Bank</h3><p>Many people assume that having a balance in an app means their money is sitting in their own bank account. In many cases, customer money is moved into pooled accounts at the partner bank. The fintech will have its own ledger detailing how much it actually owes each customer or merchant.</p><p>For example, say you’re a payroll startup and have received money for next week’s salaries. The bank won’t know how much belongs to each business. It will just know there is a large amount. The only way you’ll know how much each business and employee is owed is via your ledger.</p><p>This will only work if the records are always consistent. If there is ever a discrepancy and the fintech goes down, customers will have to wait until someone can explain why.</p><p>A customer’s balance is simply a record of how much you owe them. It doesn’t actually prove the money is there.</p><h3>2. The US Case: What Happens When Records Do Not Match</h3><p>Most recently, we can see an example in the US. A middleware firm that acted as an intermediary between fintech apps and their partner banks went into Chapter 11 bankruptcy on 22 April 2024.</p><p>Users of the apps couldn’t access their funds for several weeks or months whilst banks went through the records. There was no way of quickly finding out who owned what.</p><h3>The gap between bank balances and the ledger</h3><p>In June 2024, the bankruptcy trustee estimated a shortfall of $65 million to $96 million between the amount of cash in the banks’ pooled “for benefit of” accounts and the records held internally by the firm. The trustee also warned that there may not be a way to fully reconcile down to the last dollar.</p><p>From a freelancer’s perspective, she could see payments coming into her wallet and knew how much she had. However, there was no guarantee it was all there.</p><h3>What the regulators did next</h3><p>The US consumer finance regulator, the CFPB, took action against the firm on 21 August 2025, stating that it hadn’t kept accurate records of where consumers’ money was, or ensured that they matched the records held by the banks. It put the shortfall at $60 million to $90 million.</p><p>A judgment was entered in court on 12 September 2025 with a $1 penalty, which let victims be paid from the regulator’s Civil Penalty Fund. About $46.2 million was set aside for them in November 2025 and approximately $9 million more in May 2026.</p><p>On 2 October 2024 the FDIC published a proposal for banks that offer pooled custodial accounts. It would require banks to record the details of each beneficial owner and their balance, reconciling at least daily. It would also require banks to have “direct, continuous, and unrestricted access” to records held by a third party, even if the third party goes into bankruptcy. As of November 2026, this is still not final.</p><p>You can outsource the tech but you can’t outsource knowing where your customers’ money is and who owns it.</p><h3>3. The UK: Safeguarding Is Only as Good as Your Reconciliation</h3><p>The UK regulator, the FCA, looked into payment and e-money firms that became insolvent between early 2018 and mid-2023. It found that on average there was a 65% shortfall between what clients were owed and what had been safeguarded. This rose to 80% for the e-money firms alone.</p><p>Where money was paid back, it took an average of 2.3 years. This is the figure that really sticks in my head. Two years can cause issues for small businesses, who may miss paying suppliers and employees.</p><p>In 2023 the FCA also opened supervisory cases on about 15% of firms that safeguard. Clearly a paper plan was not good enough.</p><h3>What changed on 7 May 2026</h3><p>The FCA have introduced more stringent requirements in relation to safeguarding, which came into force on 7 May 2026. This includes:</p><ul><li>Internal and external safeguarding reconciliations on every reconciliation day.</li><li>A monthly safeguarding report to the FCA within 15 business days of the end of the month.</li><li>Annual safeguarding audits.</li><li>Resolution pack: this means you have a file of documents which can be accessed within 48 hours if the firm goes out of business.</li></ul><p>This is great. If you are a payment firm servicing a small marketplace and insolvency occurs, the clean-up team will not be searching for balances and bank statements for weeks! You should have everything in order.</p><p>Safeguarding should be a daily process.</p><h3>4. The EU: Faster and Stronger Protection</h3><p>The EU are updating their legislation with PSD3 and the new Payment Services Regulation. There was a provisional agreement reached on 27 November 2025, and the Council adopted its first-reading position on 29 October 2026. Final adoption is yet to happen.</p><p>PSD3 includes bringing e-money institutions into the payment institution regime and adding stricter requirements around safeguarding.</p><h3>From up to five days to D+1</h3><p>E-money funds must be safeguarded by the end of the business day after they are received, often referred to as D+1. This used to be up to 5 days.</p><p>If you run a travel wallet where you hold money for future trips, the longer you are holding onto funds the more exposure you have.</p><p>Your safeguarding bank must also be authorised in an EU member state. Firms should also look to avoid keeping all client funds in one bank. It would be risky to have all your merchants’ funds with one partner.</p><p>It’s about speed to safeguard and reliance on one bank.</p><h3>5. India: Merchant Money Must Stay Separate</h3><p>On 15 September 2025, the Reserve Bank of India released directions stating that non-banking payment aggregators need to maintain all funds collected for merchants in a separate escrow account with a scheduled commercial bank.</p><p>At the end of each day, they have to maintain an amount equal to or more than the outstanding amount payable to merchants in their escrow account. In other words, your escrow account must have enough funds to pay any outstanding amount due to the merchants.</p><p>Say you are a small marketplace and you collect payments from customers, and pay the seller only next week. The funds will clearly not be your working capital since they belong to someone else. This brings clarity.</p><h3>Independent checks and capital</h3><p>They also have to present auditor certificates to show escrow balances every quarter, and maintain a minimum net worth (starting with Rs 15 crore, and increasing to Rs 25 crore by the end of year 3). You can’t just rely on your dashboard.</p><p>Funds must be clearly identified as merchant funds. This is ensured with the correct banking structure in place, and independent verification.</p><h3>6. What This Means for Anyone Building a Payment Business</h3><p>It will vary depending on your country, but could you say right now how much customer money you hold, where it sits and whose it is?</p><p>Embedded finance means that you have multiple parties involved. The app may be created by one company, the banking account held by another and then the ledger held by a third party. But your customers will only ever see one app, one balance and one promise.</p><h3>What to Check Before You Hold Other People’s Money</h3><ul><li>Know where your money is stored: which bank accounts is it actually sitting in?</li><li>Know who owns the balances: keep records you’d trust to match each customer to their money.</li><li>Ensure that you reconcile daily so your ledger and your bank’s records match. This means any differences can be quickly identified.</li><li>Check that you can keep your records if your partner goes out of business. Don’t let the data live only in their system.</li><li>Ensure you know how much customer money you have in each bank, as too much in one place is a risk.</li><li>Ensure you have records in place so you can prove balances quickly.</li><li>Create external controls: audits and regulator reports should be routine, not a yearly panic.</li></ul><p>A good question to ask yourself is if you had to close tomorrow, could an external party identify all your customers’ balances and money? If you aren’t sure then there is some work to be done!</p><h3>The Real Test Comes When Things Go Wrong</h3><p>Customers will never consider pooled accounts or partner banks when using a payment app. They simply look at their balance, believe it is theirs and expect to receive it. Which is fair.</p><p>However, as seen in the US failure, it can be very painful if you don’t have matching records. As the UK figures also show, even with money safeguarded it can take years to get the money back.</p><p>Safeguarding money isn’t a policy saying the money is safe. It means you can prove it every day and return it if the business can no longer operate.</p><p>Moving money fast is easy, but safeguarding it when things go wrong is the real challenge.</p><p><strong>If your payment business was to close tomorrow, could you prove where all your customers’ money was and ensure they get it back?</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=723d9bede052" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI vs. Financial Fraud: Can Banks Stay One Step Ahead?]]></title>
            <link>https://medium.com/womenintechnology/ai-vs-financial-fraud-can-banks-stay-one-step-ahead-c4c57fe34d7c?source=rss-f965785d0636------2</link>
            <guid isPermaLink="false">https://medium.com/p/c4c57fe34d7c</guid>
            <category><![CDATA[banking]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[money]]></category>
            <dc:creator><![CDATA[Indu Pathak]]></dc:creator>
            <pubDate>Mon, 02 Nov 2026 09:46:01 GMT</pubDate>
            <atom:updated>2026-11-02T09:46:01.746Z</atom:updated>
            <content:encoded><![CDATA[<h4>The Growing Battle Between Smart Technology and Smarter Criminals</h4><p>Imagine receiving a call from someone who sounds exactly like your bank manager. They know your name, the bank you use, and even the details of a recent transaction. The caller warns you about suspicious activity and asks you to verify your account before your money disappears.</p><p>Everything sounds convincing. You follow the instructions, share a verification code, and end the call feeling relieved.</p><p>A few minutes later, you discover that your money is gone.</p><p>This is the uncomfortable reality of financial fraud today. Criminals no longer need to break into a bank or steal a physical card to access someone’s money. With artificial intelligence (AI), they can create convincing messages, imitate voices, manipulate identities, and target victims on a much larger scale.</p><p>Banks are fighting back with the same technology. AI systems can analyse millions of transactions, recognise unusual behaviour, and flag suspicious payments in seconds.</p><p>But this raises an important question: <strong>If banks are becoming smarter, why is financial fraud still such a serious problem?</strong></p><p>The answer lies in a simple truth. Technology is changing the way fraud happens, but it is also changing how criminals operate. Staying ahead requires more than sophisticated software. It requires good judgment, reliable security systems, and a better understanding of human behaviour.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-S0zyCpx7bk4IJtTe425MA.png" /><figcaption>Image is generated by AI</figcaption></figure><h3>How AI Is Changing the Face of Financial Fraud</h3><p>Financial fraud has always existed, from forged cheques to stolen credit cards. What has changed is the speed, scale, and sophistication of modern attacks.</p><p>In the past, a scammer might send poorly written emails to thousands of people and hope a few would respond. Today, AI tools can produce convincing messages in different languages, personalise them using publicly available information, and help criminals reach potential victims more effectively.</p><h3>1. Deepfakes and Identity Impersonation</h3><p>One of the most concerning developments is the use of AI-generated voices, images, and videos.</p><p>A fraudster can imitate a business executive, a family member, or a trusted financial professional. In a corporate setting, an employee might receive what appears to be a genuine video call from a senior executive requesting an urgent transfer.</p><p>The request may look completely legitimate. The voice sounds familiar, the face appears real, and the situation seems believable.</p><p>This makes traditional trust-based verification less reliable. Recognising someone’s voice or seeing their face on a screen is no longer enough to establish their identity.</p><p>Banks and businesses need stronger verification procedures, particularly for high-value transfers and changes to account details.</p><h3>2. Phishing That Feels Personal</h3><p>Most people have learned to be cautious about suspicious emails. However, AI is making phishing messages harder to recognise.</p><p>Criminals can use AI to write messages that sound professional, imitate the communication style of real organisations, and create believable explanations for why someone needs to click a link or provide sensitive information.</p><p>A message claiming that a bank account will be suspended may create just enough urgency to make a customer act without thinking.</p><p>The danger is not simply that these messages look more convincing. It is that they can be produced and adapted quickly, allowing fraudsters to test different approaches and target people more precisely.</p><h3>3. Synthetic Identities and Account Fraud</h3><p>Another growing challenge is synthetic identity fraud, in which criminals combine real and fabricated information to create an identity that appears genuine.</p><p>For example, a fraudster might use a legitimate person’s details alongside a false name, address, or other invented information to build a financial profile.</p><p>AI can help criminals generate convincing documents, alter images, and identify weaknesses in digital onboarding processes.</p><p>For banks offering online account opening, digital wallets, and instant payments, this creates a difficult balancing act: customers expect a quick and convenient experience, but identity checks must remain reliable.</p><h3>How Banks Are Using AI to Fight Back</h3><p>While criminals are exploiting new technology, banks are not standing still. Many financial institutions already use machine learning, behavioural analytics, and automated monitoring to detect suspicious activity.</p><p>The advantage of AI is its ability to examine patterns that would be difficult for a human team to identify manually.</p><h3>1. Detecting Suspicious Transactions in Real Time</h3><p>Traditional fraud detection often relies on predefined rules. A transaction above a certain amount, for example, might trigger an alert.</p><p>Rules are useful, but they have limitations. A legitimate customer might make an unusually large payment, while a criminal could deliberately keep transactions below a threshold.</p><p>AI can examine several factors together, including transaction size, payment frequency, location, device information, account history, and the recipient’s behaviour.</p><p>Suppose a customer normally makes small domestic payments but suddenly initiates several transfers to unfamiliar accounts within minutes. The system can identify this pattern as unusual and request additional verification.</p><p>The goal is not to block every unfamiliar transaction. It is to recognise when several warning signs appear together.</p><h3>2. Understanding Customer Behaviour</h3><p>Every customer develops certain financial habits. Some regularly pay bills at the end of the month, while others make frequent small purchases or transfer money between their own accounts.</p><p>AI-powered behavioural analysis can learn these patterns and identify activity that differs significantly from normal behaviour.</p><p>For instance, a transaction might be suspicious not because of its amount, but because it follows an unusual sequence of account access, device changes, and payment requests.</p><p>This approach can help banks identify fraud earlier, sometimes before a customer realises anything is wrong.</p><p>However, behavioural patterns should be treated as signals rather than proof. People travel, change jobs, buy new devices, and experience unexpected expenses. Effective systems must account for these ordinary changes.</p><h3>3. Strengthening Digital Identity Verification</h3><p>Digital banking has made opening accounts and accessing financial services easier than ever. Customers can complete many processes without visiting a branch.</p><p>AI can support identity verification by analysing identity documents, detecting possible alterations, checking facial consistency, and identifying unusual onboarding patterns.</p><p>Combined with document verification, liveness detection, and other security controls, these systems can make it more difficult for criminals to use stolen or fabricated identities.</p><p>Still, no single verification method is foolproof. Banks need layered checks that can adapt to the level of risk involved.</p><p>A low-risk activity may require little friction, while a suspicious account application or high-value transfer may justify additional verification.</p><h3>4. Connecting the Dots Across Financial Activity</h3><p>Fraud rarely follows one predictable pattern. Criminals may use several accounts, devices, payment channels, or financial institutions to move money.</p><p>AI can help investigators identify connections between seemingly unrelated events.</p><p>For example, multiple accounts may share device characteristics, receive funds from the same suspicious source, or transfer money through a similar sequence of transactions.</p><p>When these signals are analysed together, a bank may uncover a wider fraud network rather than treating each transaction as an isolated incident.</p><p>This is especially valuable in a financial environment where payments can move across platforms and borders within seconds.</p><h3>Why Banks Still Struggle to Stay Ahead</h3><p>If AI can process enormous amounts of data and identify suspicious patterns, why can’t banks eliminate fraud altogether?</p><p>Because detecting fraud is not simply a technical problem. It is also a problem of uncertainty, speed, and human behaviour.</p><h3>1. Criminals Adapt Quickly</h3><p>Fraudsters do not need to defeat every security system. They only need to find a weakness they can exploit.</p><p>When banks introduce new detection methods, criminals may change transaction patterns, use different devices, or focus on social engineering instead of directly attacking banking systems.</p><p>This creates an ongoing cycle of adaptation. Banks improve their controls, criminals change their tactics, and financial institutions must adjust again.</p><p>AI gives both sides new capabilities, so its presence alone does not guarantee an advantage for defenders.</p><h3>2. False Alarms Can Frustrate Customers</h3><p>Imagine being on holiday and having your card declined because your bank detects an unfamiliar location. Or imagine a legitimate transfer being delayed while you wait for additional verification.</p><p>These situations can be frustrating, especially when customers need access to their money urgently.</p><p>If a bank flags too many genuine transactions, customers may lose confidence in its security systems. If it flags too few, suspicious activity may go unnoticed.</p><p>Finding the right balance is one of the hardest parts of fraud prevention.</p><p>Banks must consider not only how accurately a model identifies suspicious activity, but also how it affects customers in everyday situations.</p><h3>3. Poor Data Can Lead to Poor Decisions</h3><p>AI systems depend on the information they receive. Incomplete transaction records, inconsistent customer data, and outdated fraud patterns can weaken their performance.</p><p>A model trained on historical examples may also struggle when criminals introduce a completely new technique.</p><p>Financial institutions need reliable data, regular model testing, and human oversight to identify weaknesses before they cause significant harm.</p><p>AI should support better decisions, not become an unquestioned authority.</p><h3>4. The Human Factor Remains Difficult to Predict</h3><p>Some of the most damaging scams do not involve breaking through a bank’s technical defences.</p><p>Instead, criminals persuade customers to approve payments themselves.</p><p>A victim may believe they are transferring money to a safe account, paying a legitimate supplier, or helping a relative in an emergency. From the bank’s perspective, the transaction may appear authorised.</p><p>This is where fraud prevention becomes more complicated. A technically valid payment can still be the result of deception.</p><p>Banks need to consider the circumstances surrounding a transaction, provide timely warnings, and make it easier for customers to question unusual requests before sending money.</p><h3>The Importance of Human Judgment in an AI-Driven Banking System</h3><p>There is a temptation to treat AI as a complete solution. After all, machines can analyse data faster than people and operate continuously without fatigue.</p><p>But speed does not automatically produce good judgment.</p><p>A fraud detection system may identify a suspicious pattern, yet a trained investigator may need to understand the wider context. A customer might be making an unusual payment because of a medical emergency, a property purchase, or a legitimate business transaction.</p><p>Human oversight helps distinguish between genuine risk and unusual but reasonable behaviour.</p><p>This does not mean every alert should be reviewed manually. That would be inefficient and impractical for large banks. Instead, institutions should combine automation with carefully designed escalation processes.</p><p>AI can handle routine monitoring, prioritise alerts, and identify relationships between transactions. Human specialists can focus on complicated cases, disputed decisions, and emerging fraud patterns.</p><p>The strongest approach is not AI replacing people. It is AI helping people make faster, better-informed decisions.</p><h3>What Banks Must Do to Stay One Step Ahead</h3><p>There is no single tool that can prevent every type of financial crime. Banks need a broader strategy that combines technology, operational discipline, and customer awareness.</p><h3>Invest in Adaptive Fraud Detection</h3><p>Fraud models must be monitored and updated as criminal tactics change. Banks should test their systems against emerging attack scenarios rather than relying entirely on historical data.</p><h3>Strengthen Security Beyond Passwords</h3><p>Multi-factor authentication, device intelligence, transaction confirmation, and risk-based verification can provide additional protection. Sensitive actions should receive stronger checks, especially when account details or payment destinations change.</p><h3>Make Fraud Prevention Easier for Customers</h3><p>Security measures should be clear and practical. Customers need straightforward ways to verify suspicious requests, report scams, freeze compromised cards, and contact their bank quickly.</p><p>Warnings are most effective when they explain the specific risk rather than simply telling people to be careful.</p><h3>Improve Collaboration Across the Financial Industry</h3><p>Fraud often crosses institutional and national boundaries. Where permitted by law, banks, payment providers, regulators, and law enforcement agencies should share relevant threat intelligence and develop faster ways to respond to emerging attacks.</p><p>Better coordination can help institutions recognise patterns that would remain invisible when each organisation examines its own data in isolation.</p><h3>Build Responsible AI Systems</h3><p>Fraud prevention models should be tested for accuracy, bias, privacy risks, and unintended consequences. Banks should understand why their systems generate alerts and ensure customers have appropriate ways to challenge significant decisions.</p><p>A system that catches more fraud but unfairly blocks legitimate customers is not a complete success.</p><h3>What Customers Can Do to Protect Themselves</h3><p>Banks carry significant responsibility for securing financial systems, but customers also have a role to play.</p><p>A few practical habits can reduce exposure to common scams:</p><ul><li>Never share passwords, PINs, or one-time verification codes with someone who contacts you unexpectedly.</li><li>Verify urgent payment requests through a separate, trusted communication channel.</li><li>Avoid clicking links in unexpected messages claiming to come from your bank.</li><li>Enable transaction notifications and review account activity regularly.</li><li>Contact your bank immediately if you notice an unfamiliar transaction or suspect that your credentials have been compromised.</li></ul><p>Most importantly, pause when a message creates intense urgency. Scammers often try to make people act before they have time to think.</p><p>Even sophisticated technology cannot fully protect someone who is deliberately pressured into authorising a payment without understanding what they are doing.</p><h3>The Future of AI and Financial Fraud</h3><p>The relationship between AI and financial fraud will continue to evolve. Banks will develop better ways to identify suspicious transactions, verify identities, and investigate complex networks. Criminals will continue looking for opportunities to exploit new technologies and human weaknesses.</p><p>The next stage of fraud prevention will likely involve more integrated systems that connect transaction monitoring, identity verification, behavioural analysis, and real-time risk assessment.</p><p>But progress should not be measured only by the number of suspicious transactions a bank detects. It should also be measured by how quickly victims receive help, how often legitimate customers experience unnecessary disruption, and how effectively financial institutions prevent losses before they happen.</p><p>Trust remains the foundation of banking. Customers expect their money to be safe, their information to be protected, and their access to financial services to remain reliable.</p><p>AI can help banks protect that trust, but it cannot maintain it on its own.</p><h3>Final Thoughts: Staying Ahead Is a Continuous Process</h3><p>Can banks stay one step ahead of financial fraud?</p><p>The realistic answer is that they can improve their chances, but they cannot assume the battle will ever be completely won.</p><p>AI gives financial institutions powerful tools to recognise patterns, investigate suspicious activity, and respond at a speed that would be impossible through manual processes alone. At the same time, fraudsters are using technology to make their attacks more convincing and easier to scale.</p><p>The difference will come from how banks use these capabilities.</p><p>Institutions that combine adaptive technology, strong verification, responsible data practices, skilled investigators, and customer education will be better positioned to respond to emerging threats.</p><p>Ultimately, the goal is not to build a banking system that never encounters fraud. It is to build one that can recognise danger sooner, respond more effectively, and recover quickly when something goes wrong.</p><p>In a world where criminals are becoming smarter, the banks most likely to succeed will be those that never stop learning.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c4c57fe34d7c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/womenintechnology/ai-vs-financial-fraud-can-banks-stay-one-step-ahead-c4c57fe34d7c">AI vs. Financial Fraud: Can Banks Stay One Step Ahead?</a> was originally published in <a href="https://medium.com/womenintechnology">Women in Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>