Longterm Wiki

Evals-Based Deployment Gates

in-effectInternational
Introduced 2023Wiki article →

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.

Introduced
Committee
Floor Vote
Passed
Enacted

Related Legislation

NameStatus
EU AI Actin-effect

Related Topics

Related Pages

Top Related Pages

Organizations

OpenAIUS AI Safety InstituteApollo ResearchAlignment Research CenterUK AI Safety Institute

Risks

AI Capability Sandbagging

Approaches

Red TeamingThird-Party Model AuditingStructured Access / API-Only

Analysis

AI Safety Intervention Effectiveness Matrix

Safety Research

AI Evaluations

Policy

Responsible Scaling PoliciesCompute ThresholdsNew York RAISE Act

Concepts

Alignment Policy OverviewSituational AwarenessCapability Evaluations

Historical

Mainstream Era

Key Debates

Technical AI Safety Research

Other

Beth Barnes

Quick Facts

Introduced
2023
Status
Active; EU binding, UK/Lab voluntary
Scope
International

Tags

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