Post by UCL Department of Science, Technology, Engineering and Public Policy (STEaPP)

What happens when AI can write the perfect grant… and review it too? In a new Nature comment, James Wilsdon and Geraint Rees argue that agentic AI isn’t just speeding things up, it’s challenging the foundations of how research funding works. If proposals and peer reviews are both shaped by AI trained on past successes, funding risks becoming a loop of imitation rather than innovation. So what does a funding system built for an AI-native world look like? You can find out more here: https://lnkd.in/eR6p8DTi #UCL #Innovation #AI #Research