Life Sciences Research Support
How is EU grant competition evolving these days?
We all know that grant competition is increasing. But what do the statistics tell us?
14 November, 2026
As the September-October wave of EU grant application deadlines passed, I decided to do some research to find out more about how EU funding competition has evolved since Large Language Models (LLMs, otherwise nicknamed AI) entered the market.
Most people know that LLMs have disrupted the work of many people in a lot of different sectors. Grant writing hasn’t been spared. Rumors suggest that dedicated grant writers, as well as academics, nonprofits, and companies, now make ample use of LLMs in one way or another when they’re preparing grant proposals.
The EU Commission and other funding agencies have also acknowledged that LLM-assisted writing is now commonplace. That’s why, for instance, many EU-level grant guidelines now have LLM usage instructions integrated within their application manuals. In many cases, they request an explicit declaration of LLM use. They might also warn applicants to proofread their proposals very carefully to make sure that the ideas and arguments they’re putting forth really reflect what they’re trying to propose to the evaluators.
However, rumors and media reports about LLMs only show the general direction society and workplaces may be going. When we look at the underlying numbers driving these trends, we get to understand how and why grant competition is evolving.
The numbers: more proposals per topic
At the IEEE DCOSS-IoT conference in June 2026, Texeira and coworkers presented their research study, in which they examined 2,136 EU Research and Innovation Action (RIA), Innovation Action (IA), and Coordination and Support Action (CSA) topics from 2021 to 2026. They found out that the average number of proposals per topic has increased by 136% across these Horizon Europe topics.
A 2026 Joint Research Centre (JRC) report also analyzed Horizon Europe grant applications, although this particular study focused specifically on those involving private firms. Proposals showing signs of LLM-assisted writing stayed at around 5% until November 2022. Then, after the introduction of ChatGPT in November 2022, the rate of LLM-assisted writing increased progressively to about 40% by November 2024.
These studies show that competition for grants and LLM use have both increased over the last few years. However, correlations don’t necessarily mean causation. In other words, these trends don’t tell us the extent to which LLM usage is responsible for the increase in grant competition.
EIC Transition calls and their raw public statistics
If we look at the EIC Transition calls as just one of many examples, public statistics between 2022 and 2026 also show non-negligible trends.
EIC Transition calls received:
- 143 proposals in April 2022,
- 246 in September 2022 (a total of 389 in 2022),
- 180 in April 2023,
- 257 in September 2023 (a total of 437 in 2023),
- 413 in September 2024,
- 611 in September 2025, and
- 648 in September 2026.
The EIC described both 2024 and 2025 as years when they received the record numbers of submissions. It’s true that switching from two deadlines to a single deadline per year may have had an effect on the number of applications per deadline. However, one thing for sure is that the number of applications definitely hasn’t gone down since 2024.
Does more LLM use mean more funding?
The JRC report that analyzed Horizon Europe applications involving firms also suggested that LLMs may have helped lower the entry barrier for certain groups of applicants. For example, LLM-assisted writing tended to be more common among younger firms and those from countries where people generally had lower English proficiency.
However, when applicants resubmitted proposals on the same topic across multiple years, increasing their use of LLM-assisted writing wasn’t systematically improving their odds of getting funded.
In contrast, applicants using a combination of human and LLM-assisted writing performed no worse than those using no LLMs at all.
Therefore…
These data tell us that proposal success still depends heavily on the human masterminds behind the proposals, and LLMs are best used as a complement to our creativity.
This conclusion obviously has a bit of a “duh!” factor because we assume that everybody is aware that LLMs shouldn’t write 100% of any grant application. But I’m presenting it here because there is a lot of hype, as well as a lot of fear, surrounding LLMs and their growing capabilities.
What we need to do, for now, is to rest assured. Proposal success still depends on our brains, and we need to remember to keep tinkering!
