On September 16, our CEO Gilles Meijer joined a 90-minute discussion on what happens when AI enters the evaluation of EU funding proposals.
With AI making polished proposals easier to produce, the challenge is increasingly shifting from writing to reading. More applications mean more to assess, while the important differences can become harder to spot.
A proposal can be perfectly written and still be empty in substance. Clear target groups, concrete impact and meaningful partner roles matter beyond polished language. And as AI makes it easier to tailor proposals to a call, impact and the strength of the partners become even more important.
The role of AI in evaluation is less straightforward. The discussion touched on cost, reliability and whether AI should augment rather than replace human judgement. With clear rules around AI use still developing, disclosure and transparency become important. AI may also help connect ideas that could otherwise be missed, but trust remains essential: is the information valid and verified? The track record of the organisations involved and their roles within a consortium provide important context.
At ScoutinScience, we’re working on solutions.
We use data and AI-powered tooling to help identify the right opportunities from research, find the right partners and build stronger consortia. From there, our technology helps connect research to funding opportunities and supports the development of stronger proposals.
Our proposal technology is trained on more than 6,500 reports, metadata and evaluation summary reports, giving us a unique dataset to understand what happens beyond simply writing a proposal that “fits the call.”
The goal is not to replace the people making these decisions. It is to use technology and data to help applicants and evaluators work with better information, stronger connections and more substance.
We’re working on the tools to make that possible. And we’ll be sharing more about our solutions soon.
We cannot simply step away from AI, so it is up to all of us to work together to find ways of using it that are fair, transparent and genuinely useful, while keeping human judgement where it matters.
A big thank you to the other panelists Andrea Igazi-Eppacher, Thomas Byrnes, Gilberto Martinez, Ena Hodzic, and Matteo Feruglio for sharing their experience and perspectives, and to Annika Jaansoo for bringing everyone together and moderating the discussion. And of course, thank you to everyone who joined and contributed to the conversation.
