Can AI coding agents rebuild biological AI tools from a paper?
AISI's ReproBAIT evaluation finds coding agents can rebuild simple biological AI tools from limited public info within 4 hours and under $20, undermining withheld-code access controls.
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 Evals & standards
Date Any time
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.
How Far Behind the Frontier are Leading Open Weight Models on Cyber?
AISI measures cyber capabilities of open and closed-weight models, finding leading open models now trail frontier closed models by 4-7 months, and discusses misuse risks of open release.
Deepening our partnership with the Australian AI Safety Institute
Announces an agreement between the UK and Australian AI Safety Institutes to collaborate on AI evaluation best practices and share research findings. Institutional news with no technical result.
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.
Our evaluation of OpenAI's GPT-5.5 cyber capabilities
UK AISI reports cyber evaluations of GPT-5.5, finding it is the second model to complete their multi-step corporate network attack simulation end-to-end, suggesting a broader frontier trend.
How are AI agents used? Evidence from 177,000 MCP tools
Measures 177,000 public MCP agent tools, finding action tools rose from 27% to 65% of usage, and proposes tool-layer monitoring so regulators can oversee risky agent deployments.
International consensus and open questions in AI evaluations
Summarises an international AI safety institute network's consensus on AI evaluation best practice, including clear objectives, transparency and reproducibility, and open measurement questions for frontier AI.
Our 2025 year in review
UK AISI's year-in-review highlights its first Frontier AI Trends Report, which tracks dangerous capability growth in cyber, bio, and chemistry from two years of testing, plus government preparedness work.
AISI Frontier AI Trends Report (2025)
UK AISI's first public report synthesising two years of frontier model testing, tracking trends in dangerous capabilities across national-security and public-safety domains for policymakers and the public.
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.
Open technical problems in open-weight AI model risk management
Lists 16 open technical challenges for managing risks of open-weight frontier models across training data, algorithms, evaluations, deployment, and ecosystem monitoring, noting limited safety tooling.
Navigating the uncharted: Building societal resilience to frontier AI
UK AISI's Societal Resilience Team outlines how it will monitor real-world deployment of frontier AI and provide evidence-based risk assessments to government, covering fraud, self-harm, finance and infrastructure harms.
An example safety case for safeguards against misuse
Proposes an end-to-end safety case showing that misuse safeguards reduce AI assistant misuse risk to low levels, combining red-teaming effort estimates with a quantitative uplift model.
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.
Research Agenda
UK AI Security Institute sets out its research priorities, its approach to technical solutions for pressing AI concerns, and the key risks to address as capabilities advance.
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.
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.
Pre-Deployment evaluation of OpenAI’s o1 model
UK and US AI Safety Institutes jointly ran pre-deployment tests of OpenAI's o1 across cyber, biological, and AI-development capabilities, comparing it against reference models and sharing findings before release.
Safety case template for frontier AI: A cyber inability argument
Proposes a safety case template using CAE structured arguments to show frontier AI models lack unacceptable offensive cyber capabilities, linking risk models, proxy tasks and evaluation results.
Safety case template for ‘inability’ arguments
Proposes a safety case template for arguing that an AI system lacks offensive cyber capabilities, combining existing techniques as a proof of concept for frontier AI assurance.
Our First Year
UK AI Security Institute's first-year reflection on building government capacity to run dangerous-capability evaluations and shape an international approach to measuring frontier AI risks.
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.
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.
Fourth progress report
UK AI Security Institute's fourth progress report covering its first model evaluation blog post, the International Scientific Report on AI safety, Inspect release, office and partnership news, and staff growth.
Announcing the UK and US AISI partnership
The UK and US AI Safety Institutes agree to jointly test advanced AI models, share research, exchange model access, and transfer expert talent.
Third progress report
UK AISI progress report on recruiting researchers, publishing principles for the International AI Safety Report, and starting pre-deployment testing of advanced AI systems. Mostly organisational news.