AIRiskAware
AI Governance Glossary
Governance Concept

What Is Frontier AI Safety Commitments?

Frontier AI Safety Commitments is voluntary commitments by leading AI companies to assess, manage, and publish their approach to severe AI risks.

Definition

Frontier AI Safety Commitmentsvoluntary commitments by leading AI companies to assess, manage, and publish their approach to severe AI risks.

Announced at the 2024 Seoul AI Safety Summit, these commitments saw a group of major AI companies agree to publish safety frameworks, set risk thresholds at which they would not deploy a model, and be transparent about their approach. They are voluntary, but they established a shared baseline expectation for how frontier developers should handle severe risks.

Source: Seoul AI Safety Summit (2024)

Plain-language explanation

Announced at the 2024 Seoul AI Safety Summit, these commitments saw a group of major AI companies agree to publish safety frameworks, set risk thresholds at which they would not deploy a model, and be transparent about their approach. They are voluntary, but they established a shared baseline expectation for how frontier developers should handle severe risks.

Primary source: Seoul AI Safety Summit (2024)

Related terms

AI Safety Frontier Model Responsible Scaling Policy (RSP) Bletchley Declaration

See where you stand on AI governance

Take the free 7-question maturity assessment and get a personalised action plan.

Free assessment — 3 minutes →