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.
Response plans, escalation procedures, and incident management for operational emergencies.
Also in Operations & Security
Stage: Escalation; Stakeholder: AI Developers; Additional information: AI developers should establish well-defined escalation protocols and conduct regular training exercises to ensure their effectiveness. Developers should create incident response plans in advance, with well-defined, evidence-based thresholds for when to trigger an emergency response. Incident plans should assign critical roles, including an 'incident commander' who has decision making authority, direct access to leadership and the authority to coordinate cross-functional teams and suspend models. Incident protocols should be customisable to accommodate variations, and organisations should drill escalation pathways (Webb & Chevreau 2006).
Reasoning
Regular training and scenario drills prepare teams to execute incident management and escalation procedures.
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.
Operate and Monitor
Running, maintaining, and monitoring the AI system post-deployment
Developer
Entity that creates, trains, or modifies the AI system
Manage
Prioritising, responding to, and mitigating AI risks
