Don’t Ghost Grants: How New AI can Supercharge your Effort
By Jacob Borgeson and David J. Neff
Most founders Jacob and I talk to have left six figures of non-dilutive funding on the table because they thought grants were too slow, bureaucratic, and painful. They were right in 2019. They’re wrong in 2026. What used to take months of work now takes days or weeks. Never hours.
All of this work is now being simplified by the ultimate equalizer of 2026, AI. That’s what this article is all about. How can you find, apply for, win and then manage grants for your startup using AI.
Grants are bigger than you think and can open doors
There are >$400B in R&D grants and roughly $40B in commercialization grants awarded every year. SBIR maybe the most famous, but it’s just $4B per year. And many grants are not nearly as arduous as SBIR. Although government grants declined globally by 5% in 2025 (including a 7.9% decrease in the US driven by federal funding disruptions), it is still a big pool of money worth considering.
Let’s be clear about something — this government decline is not good policy. And the 7.9% US decline isn’t noise, it’s signal. Federal grant programs, especially DOE and NSF, are operating in a more volatile political environment than at any point in the last 20 years. It’s a huge drag on innovation.
CALLOUT: Grants also increase your odds of winning VC rounds: ARPA-E data shows that winning an early-stage grant roughly doubles the probability of securing subsequent VC funding.
The variety of grants is enormous and expanding
To be clear, you shouldn’t abandon Federal or SBIR grants. Your grant strategy in 2026 needs geographic and institutional diversification the same way your investor list does. The sheer variety of private foundation grants, international programs, and state-level programs are underutilized by the US deep tech founders we work with. And the size of these non-government pools is expected to grow and focus more on innovation and commercialization, with some estimating the pool of reserve funds for private foundations growing from $2T today to $18T by 2050.
How to get more grant money today
There are a ton of new tools to take advantage of grants. And like the founders we interviewed, many of you have probably tried Claude or ChatGPT. And you may have come across other software, but been hesitant to try it due to time or skepticism.
As you can see, we tagged these platforms by broad categories of capabilities to help with the big steps in the grant process.
A quick shameless ploy for interaction
What we did is a quick overview of a new market that is growing fast. We definitely missed some, didn’t dive deeply ourselves into all of them, and new features are released daily.
So let us know your experience and anything we missed in the comments.
#1 FAQ of 2026: Can’t I just use the AI that I have?
You’ve already spent time rolling out and paying for Claude, or OpenAI or Gemini. And they are very helpful with almost every task. Do you need anything else? It’s a great point, so we asked Dr. Manas Kala who’s the founder of Veritus.AI
“Claude and Gemini are excellent general-purpose models, and we use them ourselves, but they optimize for generation, not for whether you can trust the output when your paper, grant, or career is on the line. As AI-generated content becomes the norm, review intelligence matters as much as generation.”
And using these tools isn’t without merit. Almost every founder we talked to had seen benefits from applying their AI assistants to win grants.
Dr. Gabriel Cossio of Nanoscale Labs says Claude has improved his process dramatically (and won a NASA grant)! His process looks like this:
“Brain dump my ideas in a semi-stream of consciousness manner into a Claude project → ask Claude to write out a draft for a section of the grant using that content → Claude gives me back a personalityless semi-robotic draft that’d I’d grade at a D+. However, reading Claude’s crappy draft helps me distill and streamline my ideas into a functional argument without spending hours and hours on brainstorming prose and outlining.”
Purpose-built tools enable diversification of grants
As you’ll see, these solutions help you be more efficient across the entire grant lifecycle — not just writing. By bringing you results from all grant types, whittling that down to the top 5 that fit you, helping you navigate bureaucratic systems, and accelerating proposal drafts means you can more easily access a broader pool of non-dilutive funding.
Search: Game Changing
No matter your opinion on AI, one of the big benefits is the ability to make sense of large sets of complex, messy data. Grants 110% fall into this bucket with 80,000+ potential sources and 3 million grants open throughout the year across the world. (Source: Candid)
The first step is to sign up for a solution and start searching — all of the solutions above make it really easy for you. You can get pretty tailored results in a few minutes as they ingest data about what you do and what you need from documents or your website. From there it becomes a chat with a dedicated AI assistant to further narrow down and refine results.
Behind the scenes are matching algorithms that constantly improve through usage and quickly provide matches from published eligibility requirements, timing, key criteria, and estimated effort level.
We asked Ray Crick, CEO at Allegory.world about this:
“Finding the right grant is like finding the right needle in multiple haystacks while also being able to tell if that needle is going to be more of a burden than it was worth. We know from experience that post-award work is more than half the battle, and most founders don’t have that kind of time. This is what many consultants get paid thousands of dollars to do. New AI models help with finding and writing, but struggle with predicting post-award effort, fit, and especially compliance. We’re building Allegory for the entire grant lifecycle because search is just the beginning.”
Registration aka Wading through Bureaucrap:
However, AI cannot do the registration process required in multiple systems for grants like SBIR. It can, however, help you understand the process, find the hidden links and sites, and help you collect and format your information to complete the registration. Which is a huge help.
TL;DR: There is a huge improvement in finding grants with AI.
A quick few minutes of search can now easily net you 3–5 grants you can apply for from multiple sources. It also helps you understand and manage the eligibility process and match with grants that aren’t a waste of time. This used to take 20+ hours of effort or an expensive consultant.
Proposal Writing: Mostly better with AI
Just the base level of importing clunky websites or PDFs into an AI-assisted workspace is a huge benefit. Instead of spending hours diving into many pages of documents that funders put out about what they are seeking to accomplish, what criteria they are using, or how they want proposals formatted — your AI-assistant can do all that for you and put together an outline of a proposal based on your knowledgebase and conversational guidance.
Get David J. Neff’s stories in your inbox
Join Medium for free to get updates from this writer.
Some, like Allegory.world even help you build a strategy, project checklist, and tasks to help you stay organized and collaborate with team members.
“Overall, I’ve probably gone from needing a month or two to develop what I consider a compelling grant application to more like a week or two.” — Dr Gabriel Cossio.
However, there are still challenges remaining.
First, for highly technical and cutting edge innovation, such as for SBIR, AI cannot write about content it hasn’t learned about. You will still not be able to rely on it to describe the technical aspects of your work.
Dr. Rawand Rasheed, founder and CEO of Helix Earth said
“What mainly is an issue with AI writing grants is that AI’s language and writing is pretty generic. When you are writing about a new product, technology, or research path, it can tend to be hyper-specific domain knowledge, and AI just is not great at writing for anything that specific.”
TL;DR — Marked improvement in proposal writing with real ROI by using AI.
Having an AI-assitant that knows you, can understand the grant documents quickly, and then collaborate with you is a benefit that not only saves weeks of effort, but can also make your proposal better. But you still need to add your expertise.
Management & Reporting: definite potential with AI
Winning a grant doesn’t mean you are done. Post-award management and reporting can take up to 42% of your time. Some of these new solutions also continue to help with this challenge. There is accounting involved, impact reporting (i.e. what you accomplished in the world with the money), and more forms to fill.
Most of these solutions haven’t been around long enough to see how much they help here. The ones that have (i.e. Instrumentl, Submittable) are working on this aspect of the problem, but they are more for non-profits and not R&D. They are also by far the priciest on the list.
TL;DR: Impact unknown today for reporting with AI
We’ll have to wait and see in a year or two after these platforms have been in use more.
What about Costs?
The cost for most of these solutions is dramatically lower than the old way of doing things. Engaging a grant consultant used to start at thousands of dollars per month, and could include a % of the overall award as a success fee. This was a hard stop to most startups we know of.
On the low end, you can benefit from tools you already use like Claude or ChatGPT. And some of the more powerful purpose-built tools we reviewed now include a free trial, freemium version, and pricing from $50 to $1500 per month depending on number of users and features.
TL;DR: Definitely cheaper than hiring a person, but you still have to put work in.
Conclusion
Welcome article skimmers! This is a quick summary of an article that got way longer than we thought it would!
Not everything is coming up roses: easier applications = more competition.
As with most areas where generative AI has enabled more people to communicate better, easier writing = more people writing. Grantmakers are also seeing more applications than ever and they have their own AI to detect and remove generic applications.
This means making your application or proposal stand out is more difficult.
What AI can’t do is build the relationship with the program officer, get you the warm intro to the foundation director, or tell you which grants are actually worth winning versus which ones will bury your team in compliance for 18 months. That’s still the hard part. That’s still the human part.
And in the age of AI, it’s what sets us apart.
Our final advice?
Go sign up for a trial with one of the software vendors that we cover in this article and see what you can do with just one opportunity over the course of a week or a month. We guarantee you will find at least one grant that fits you that you had not heard of before.
- Cut the search time from a week to an afternoon.
- Cut the drafting time from a month to a day.
- Ease the whole process dramatically with a dedicated, tuned AI assistant.
- Improve efficiency to apply for MORE GRANTS through collaboration and task management.
And come back and tell us how it went in the comments below.
David J. Neff and Jacob Borgeson have spent a combined 35+ years helping founders navigate the gap between breakthrough technology and the capital to build it.
David runs Neon Syndicate Advisory in Texas and NY, and works across Deep Tech and Energy Startups. Jacob is an independent advisor working with Climate Tech and AI startups in Texas and Europe. Jacob advises Allegory; but all opinions are his own.
