Led by Surrey, the three universities will design a digital sandbox that provide a secure, tightly controlled environment for AI agents.
The secure sandbox project has received a grant of more than £3.2 million from the Advanced Research + Invention Agency (ARIA) as part of its Safeguarded AI programme.
AI agents are becoming capable of carrying out increasingly complex tasks independently, with recent incidents involving Anthropic and OpenAI highlighting the need for better safeguards that keep them within defined boundaries. The Agentic-seL4 project will explore how AI agents can be safely contained.
Led by Research and Development Creator Professor Brijesh Dongol from Surrey’s Centre for Cyber Security, the team will build a rigorously verified sandbox that controls what AI agents can access and do, with permissions that can be changed or removed while they are operating.
For example, an AI agent that has accessed confidential information could be blocked from communicating externally or allowed to do so only after its output has passed through a trusted sanitisation process.
The system will be backed by formal models and proofs that guarantee the specified policies are always enforced, giving organisations mathematical assurance that sensitive data cannot leak. The project will also develop a proof-of-concept reference design, which will be tested against realistic “agent-escape” scenarios to validate how well these restrictions hold.
Providing freedom for AI agents to problem-solve, whilst staying in safe limits
Professor Brijesh Dongol, Director of the UK Research Institute on Verified Trustworthy Software Systems (VeTSS), commented: “There’s no denying that AI is advancing at an extraordinary pace, and the safeguards we put around these systems need to advance with it.
“Our aim is to give AI agents the freedom to carry out useful tasks, but within boundaries we can trust. Providing rigorous guarantees about what an agent can access and do will be an important step towards safely using AI, particularly in security-critical environments.
Bringing together researchers from the Universities of Surrey, Melbourne, and Sheffield, Agentic-seL4 is one of eight new research projects that have received a share of £22 million through the TA2 Cybersecurity track of ARIA’s Safeguarded AI programme, led by Programme Director Nora Ammann.
The programme is exploring whether AI-enabled formal methods – mathematical techniques that prove software and hardware behave as specified – can make high-assurance cyber defence practical at a scale previously out of reach.
