Longterm Wiki

Evals-Based Deployment Gates

Evals-based deployment gates require AI models to pass safety evaluations before deployment or capability scaling. The EU AI Act mandates conformity assessments for high-risk systems with fines up to EUR 35M or 7% global turnover, while UK AISI has evaluated 30+ frontier models.

Related

Related Wiki Pages

Top Related Pages

Risks

AI Capability Sandbagging

Analysis

AI Safety Intervention Effectiveness Matrix

Approaches

AI EvaluationThird-Party Model AuditingStructured Access / API-OnlyDangerous Capability Evaluations

Organizations

OpenAIUS AI Safety InstituteApollo ResearchAlignment Research CenterUK AI Safety Institute

Concepts

Alignment Policy OverviewCompute ThresholdsSituational AwarenessCapability Evaluations

Policy

New York RAISE Act

Other

Red TeamingBeth Barnes

Historical

Mainstream Era

Key Debates

Technical AI Safety Research

Tags

evaluationsdeployment-gateseu-ai-actsafety-testingthird-party-audits