This page is still being polished. If you have thoughts, please share them via the feedback form.
Data on this page is preliminary and may change. Please do not share or cite these figures publicly.
Risk & Assurance mitigations not clearly fitting above categories.
Also in Risk & Assurance
Stage: Containment and Mitigation; Stakeholder: Third Party Researchers; Additional information: Governments and developers should improve safety governance by fostering robust safety cultures and adopting secure-by-design principles. AI developers should evaluate failure modes and implement safeguards before deployment, with independent third-party audits verifying compliance with existing standards, as commonplace in other fields such as nuclear energy, aviation, finance and banking, pharmaceuticals, and more. AI developers should also continue to allocate compute resources to AI safety, including research on monitoring, alignment and safeguards. Governments could make secure-by-design guidelines mandatory to ensure that safety features are built into AI models from the outset (NCSC 2023). They could also consider requirements or incentives for safety research.
Reasoning
Assessment activities span red-teaming exercises (2.2.2), evaluation procedures (2.2.2), and auditing (2.2.3) within same L2.
Monitor critical capability levels
2.2.2 Testing & EvaluationIdentify early warning signs and emergent capabilities
2.2.1 Risk AssessmentEstablish standardised benchmarks and reporting
3.2.1 Benchmarks & EvaluationImplement compute monitoring and anomaly detection
1.2.3 Monitoring & DetectionEnhance hardware and supply chain oversight
2.3.3 Monitoring & LoggingLead efforts to establish shared criteria for AI LOC
3.2.2 Technical StandardsStrengthening Emergency Preparedness and Response for AI Loss of Control Incidents
Somani, Elika; Friedman, Anjay; Wu, Henry; Lu, Marianne; Byrd, Christopher; van Soest, Henri; Zakaria, Sana (2025)
As artificial intelligence (AI) systems become increasingly embedded in essential infrastructure and services, the risks associated with unintended failures rise. Developing comprehensive emergency response protocols could help mitigate these significant risks. This report focuses on understanding and addressing AI loss of control (LOC) scenarios where human oversight fails to adequately constrain an autonomous, general-purpose AI.
Verify and Validate
Testing, evaluating, auditing, and red-teaming the AI system
Other
Actor type not captured by the standard categories
Measure
Quantifying, testing, and monitoring identified AI risks