The Internet Is Changing: What Does This Mean For Brands?
There has long been a theory that the majority of internet traffic and online posts, once attributed to humans, has been replaced by AI and bots. Coined The Dead Internet Theory in 2021, it suggests that these bots are being used to shape the thoughts and beliefs of humanity through said posts.
We have all experienced interactions with bots online, whether it’s comments on an X (Twitter) post or follow requests from a clearly fake Instagram profile, bots and automated accounts are everywhere. Now in 2026, Instagram reels is flooded by bot accounts, posting content with completely unrelated captions and descriptions, taking the place of human created content.
The theory that bots could take over the internet may be hard to believe, but in recent times, it has become much less a theory and more of a reality.
Data from Cloudflare suggests that during a week throughout early June 2026, bots accounted for 57.2% of internet traffic, humans making up the other 42.8%.
This doesn’t feel majorly surprising with the number of bot accounts both posting comments and following accounts as I mentioned earlier.
For the average person, these statistics make for no more than scary reading. It can be irritating receiving daily follow requests from clearly fake accounts and having ‘please send me this post’ commented on every Instagram share, but, on the other hand, more optimistic hand, some people may even appreciate getting an extra few likes on their holiday pics regardless of who they are from.
When Engagement Is No Longer Human
For brands however, it’s a whole other story.
Social media has become an essential tool for brands, offering them the opportunity to get closer to consumers and ensure their product are always just one scroll away.
Likes and comments help social media marketers gauge the effectiveness of their content, but if those likes and comments are coming from autonomous accounts, they create skewed data. Data which is essentially meaningless and can push social media managers into creating content that is ineffective in helping their brands grow.
With the advancements of AI, these bots are becoming extremely advanced and more and more capable of interacting with online content in a convincingly human way.
AI Is Changing How We Discover Information
These bot accounts are fairly unique to social media, but it’s not the only way in which AI is changing the internet.
AI LLMs, like ChatGPT, Gemini and Claude offer users the opportunity to get in-depth answers to (almost*) any question. I’m sure you don’t need telling, but these LLMs pull data from hundreds of sites and neatly wrap it up into one detailed response.
*I say almost, because they won’t teach you how to build a gun or create napalm in your bathtub.. and thank God for that.
Despite being autonomous and unaffected by human intervention, these LLMs certainly have their favourite resources.
According to Ahrefs.com YouTube and Reddit are the websites most referenced by Google’s AI overviews at 20.9% and 19.6% respectively, followed by Facebook at 11.6%. Though this data is gathered in the USA, so the results may vary if the research was conducted in other terrotories.
The issue here is, users tend to read the Google AI overview rather than visiting the website the information is pulled from. The quick summaries mean internet browsers are less likely to scroll down the Google results page, making SEO and fighting for the top spot on Google Searches less effective than it once was.
Get Adam’s stories in your inbox
Join Medium for free to get updates from this writer.
This is a problem for advertisers and brand managers, 74% of whom believe that AI overviews are reducing the number of clicks their site is receiving therefore, reducing traffic; a rather large issue for sites that rely on ad revenue as core income. Even if their website is the one that holds the information put forward by Google AI, because the user never actually visits the information’s source, there is no click and no ad revenue. Additionally, it makes it increasingly difficult for said websites to garner loyal visitors, after all, how likely are you to go back to a webpage you never actually visited in the first place?
The End of the Traditional Search Journey?
Beyond Google AI overviews, data from Semrush suggests that once a user has asked a question on ChatGPT, they are unlikely to follow through to the website the data and information has been scraped from.
Now, the Semrush data could be somewhat contentious, as it infers that the ChatGPT’s website traffic journey leaves OpenAI.com, parent company of ChatGPT, as the most common online journey destination for users, post ChatGPT usage.
Because of the unlikeliness of users following through to OpenAi.com, one interpretation of this data is that ChatGPT has become the final destination in many users’ information journey. They have gotten the answer to their question, so only intrigued users looking to delve deeper into the topic are likely to continue onto a referenced site. After OpenAI.com, YouTube follows, confirming that many LLMs have go-to sources, rather than being something unique to Google’s AI overviews.
Along the same vein of AI ‘visiting’ only a few sites, internet users (of the human variety) are visiting lews. Data from Ofcom found that in 2025, 40% of adults in the UK ‘never’ visit new sites, instead opting to stick with those they have become familiar with.
14% did say that they visit new sites ‘a lot’, but this number is down 10% when compared to 2024.
What Does This Mean For Brands?
This data, combined with changes in users’ online behaviour, has put brands in a difficult position. The internet is and always has been an incredibly complex and competitive space and despite billions of users, reaching real, potential customers has become increasingly unlikely.
I understand that the title may suggest some kind of foolproof solution for brands, but, I hate so say it: there isn’t one.
At the moment, the best bet for brands that want to reach human consumers is remarkably unclear, not so dissimilarly to when the internet was initially introduced.
Social media posts are ineffective if not reaching a real human audience and investing in SEO/SEM may have no impact on traffic, if even the top results are not being seen by users, then the second and third are almost certainly going to go unnoticed.
Currently, despite the issues with bot/autonomous accounts and potential ineffectiveness, social media is still a viable option for reaching an audience.
Although, it is more vital than ever that social media marketing managers closely monitor interactions and learn to differentiate between bot and human traffic to be able to truly understand exactly which type of content is and isn’t effective.
Secondly, brands must work on ensuring their product is at the centre of organic conversation on sites such as Reddit and YouTube. Ensuring real and trustworthy reviews are posted by reputable creators and normal people to increase their chances of being mentioned on AI chatbots.
At the end of the day, we are going through a period of change that is unprecedented and was unlikely to have been predicted. Because of this, a brand’s only option is to ensure that no matter what, they are offering something that delivers a genuinely positive and life improving proposition to ensure consumers keep coming back. Without this, topping the ChatGPT references and efficient, well monitored social media management mean nothing.
The Future Is Human
My final point is to us all, as consumers.
We must not allow bots and AI chatbots to determine what we see and in-turn influence how we think. We must ensure that despite the convenience of AI engines, we do not allow them to become our only sources of information.
So next time you Google something or ask ChatGPT what you should have for dinner, ignore the overview, click the links and see it for yourself.
