Towards technical best practice for frontier AI development
AISI sets out technical best practices and controls for frontier AI developers, drawing on its pre-deployment testing experience, to address risks emerging during AI development.
Research / Sources
UK AI Security Institute
The UK government’s AI Security Institute: evaluations, red-teaming and research on frontier AI risks. aisi.gov.uk
Organisation (36)Red Team (13)Science of Evaluations (11)Cyber & Autonomous Systems (10)Engineering (6)Human Influence (5)Control (4)Societal Resilience (4)Safety Cases (3)Chem-Bio (2)Model Transparency (2)Strategic Awareness (1)
Focus All
Topic Lab practice
Date Any time
Will it become harder to oversee AI systems?
Report mapping AI oversight methods (audits, monitoring, incident investigation) across internal activations, chain-of-thought, actions and inter-agent channels, and how they may degrade as capabilities advance.
Funding 60 projects to advance AI alignment research
UK AISI announces first 60 grantees of the Alignment Project, a £27m international funding programme for AI alignment research, with new partners joining the coalition.
Our approach to tackling AI-generated child sexual abuse material
UK AISI and Thorn publish a safety protocol for AI developers to prevent AI-generated CSAM, alongside legislation allowing authorised testing of whether models can generate such material.
Practical challenges of control monitoring in frontier AI deployments
Analyses design choices for control monitors of AI agents in real deployments (synchronous, semi-synchronous, asynchronous), identifies oversight, latency and recovery challenges, and uses a safety case sketch.
From bugs to bypasses: adapting vulnerability disclosure for AI safeguards
Explores whether cybersecurity vulnerability-disclosure practices can help manage bypasses of generative AI safeguards, in collaboration with the UK NCSC.
Managing risks from increasingly capable open-weight AI systems
UK AISI post on managing risks from capable open-weight models, covering why released weights let safeguards be cheaply removed, existing mitigation methods, and open problems.
Making safeguard evaluations actionable
Proposes how to make misuse-safeguard evaluations decision-relevant by setting metrics and thresholds, and presents an example safety case linking safeguard evaluation results to deployment decisions.
Safety Cases: A scalable approach to Frontier AI safety
Argues safety cases can demonstrate frontier AI safety to decision-makers, linking them to the Frontier AI Safety Commitments and listing open methodology questions.
How can safety cases be used to help with frontier AI safety?
Explains how safety cases, structured arguments linking claims and evidence, can help frontier AI developers check whether systems meet the thresholds in their safety frameworks.
Emerging practices in frontier AI safety frameworks
Summarises emerging practices for frontier AI safety frameworks across risk identification, mitigation and governance, drawing on companies, governments and researchers after the Seoul commitments.
A sketch of an AI control safety case
Sketches a "control safety case" argument that an internally deployed LLM agent cannot exfiltrate data, relying on red-team control evaluations and extrapolation claims.
Early lessons from evaluating frontier AI systems
UK AISI blog on third-party evaluation of frontier AI systems, covering which systems, risks and capabilities to test, and how to run robust pre-deployment safety evaluations.
Why I joined AISI by Geoffrey Irving
Geoffrey Irving explains why he left alignment research for the UK AI Safety Institute, arguing safety progress is too slow and technical work informing government and risk coordination is critical.
Conference on frontier AI safety frameworks
Announces an invite-only AISI and GovAI conference and call for submissions on designing and implementing frontier AI safety frameworks, following the Seoul commitments.
Safety cases at AISI
AISI announces planned research sketching safety cases, structured arguments that frontier AI systems are safe, focused on loss-of-control and autonomy risks for future models.