Founded in 1907 and headquartered in South Kensington, London, Imperial College London is ranked #2 globally (QS 2026) and conducts research across engineering, natural sciences, medicine, and business. Its AI safety-relevant work spans the UKRI Centre for Doctoral Training in Safe and Trusted AI (co-hosted with King's College London), an AI Security and Privacy Lab studying adversarial risks in AI systems, and a partnership in the Leverhulme Centre for the Future of Intelligence alongside Cambridge, Oxford, and UC Berkeley. Imperial played a prominent role in the UK AI Safety Summit 2023 and hosts multiple UKRI-funded centres for doctoral training in AI.
Funding Details
- Annual Budget
- $1,394,700,000
- Monthly Burn Rate
- $116,225,000
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- -
Theory of Change
Imperial's theory of change for AI safety operates through three main channels. First, by training a new generation of researchers in safe and trusted AI via doctoral programmes (STAI CDT and AI for Healthcare CDT), Imperial aims to build the human capital needed to develop and deploy AI systems that are verifiable, interpretable, and robust. Second, through foundational technical research in formal verification, symbolic AI, adversarial robustness, and privacy-preserving methods, Imperial contributes the theoretical and empirical foundations that future safety-critical AI systems can be built on. Third, through policy engagement and partnerships with government and industry (including the AI Safety Summit, Thomson Reuters Frontier AI Lab, and UKRI funding), Imperial bridges academia and real-world deployment to ensure safety research informs practice. The LCFI partnership addresses longer-term civilisational and existential considerations around AI development.
Grants Received
from Long-Term Future Fund
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 9:56 PM UTC
- Created
- Mar 19, 2026, 10:42 PM UTC