Universities and research institutions may need to rethink how they manage research funding and academic support services as the rise of ‘agentic AI’ threatens to overwhelm existing grant application systems, according to new analysis from UCL researchers.
Writing in Nature, Professors Geraint Rees and James Wilsdon warn that increasingly advanced AI tools are fundamentally changing how research proposals are created, raising significant implications for higher education operations, governance and resource management.
Unlike earlier AI systems that respond to simple prompts, agentic AI tools can independently complete multi-step tasks, including analysing funding criteria, reviewing past successful applications and generating fully formed grant proposals with minimal human input.
For leaders across colleges and universities, the findings highlight a broader challenge facing the education sector: managing the infrastructure, systems and operational pressures created by rapid AI adoption.
The researchers argue that funding bodies are already experiencing strain. Analysis of hundreds of thousands of applications across 12 international research funders found application volumes increased by 17% between 2022 and 2024, rising to 57% growth by 2025. Some programmes saw applications more than double.
This surge has implications far beyond research departments. Increased competition for funding is likely to place greater pressure on institutional support teams, digital infrastructure and compliance processes, while also influencing future investment priorities around AI governance, data management and research administration systems.
At the same time, universities are being urged not to simply restrict AI use, but to adapt operationally. The researchers suggest funding organisations may eventually need to deploy AI themselves to help prioritise, assess and verify applications at scale.
For estates and procurement leaders, this points to growing demand for secure digital infrastructure, enhanced computing capability and AI-ready environments that can support both academic activity and institutional oversight.
The analysis also raises wider governance concerns. Without careful implementation, AI-driven funding systems risk reinforcing existing inequalities by favouring established institutions or researchers with stronger historic funding records.
As AI capabilities continue to evolve rapidly, the report suggests higher education institutions will need to balance innovation with transparency, compliance and operational resilience, while ensuring the systems supporting research remain scalable, fair and fit for an increasingly AI-driven future.
Photo by Stephan HK on Unsplash
