AI Safety Accountability

Static Demo

Tracking AI safety promises across frontier model developers

Simulation coming soon — currently showing static accountability data.

56

Network Health

C

Grade

5

Promises

Promises

AI-001DegradedTransitional
32%

Publish model cards with safety evaluations for all frontier models before release

OAICIVTransparency+givevoluntary

Model cards published but safety evaluation detail varies significantly across releases.

AI-002VerifiedTransitional
88%

Maintain responsible scaling policy with defined capability thresholds

ANTCIVSafety+givevoluntary

Anthropic's RSP is published and updated. ASL levels defined with clear thresholds.

AI-003DeclaredTransitional
55%

Participate in pre-deployment safety testing for frontier AI models

GOONISSafety+givevoluntary

Commitment made but the testing framework is still being defined by NIST.

AI-004DeclaredTransitional
55%

Implement robust content provenance for AI-generated media

OAICIVTransparency+givevoluntary
Depends on:

C2PA metadata support announced but not yet deployed across all products.

AI-005DegradedTransitional
32%

Publish AI safety evaluation framework for frontier model assessment

NISCIVVerification+giveimposed

NIST AI RMF published but frontier-specific evaluation benchmarks remain incomplete.

Insights

Voluntary Commitments Lack Enforcement

Verification Gap

All AI safety promises are voluntary. There is no statutory enforcement mechanism comparable to Oregon HB 2021's regulatory structure.

Responsible Scaling as a Model

Working Mechanism

Anthropic's RSP represents the most structured voluntary commitment in the AI safety space, with defined capability thresholds triggering additional safeguards.

Want your commitments mapped like this?

We build interactive promise graphs for organizations, advocates, and policy teams.