Humanein the Loop
← The matrix
06

AI should have internationally agreed-upon limits

Current path

AI development races ahead of governance. Frontier labs operate with voluntary and inconsistent safety commitments. Dangerous capabilities emerge before international agreement is possible.

Better future

AI development operates within internationally agreed limits. Safety thresholds are shared across jurisdictions. Export controls and compute governance align across allies. Dangerous capabilities are coordinated, not raced.

Drift across the three domains

Norms

Advancing17 signals
CHT recommends
  • Build international scientific consensus on AI risks (IPCC-style).
  • Foster multilateral civil society coordination on AI governance demands.
Indicators we track
  • 6.N.aInternational scientific consensus on risks
  • 6.N.bMultilateral civil society coalitions
  • 6.N.cPublic expectation of cross-border governance

Laws

Advancing16 signals
CHT recommends
  • Advance binding multilateral instruments (e.g., Council of Europe AI Convention).
  • Coordinate export controls on frontier compute and models.
  • Establish compute governance and licensing regimes.
  • Codify shared safety thresholds (bioweapon uplift, cyber offense, autonomy).
Indicators we track
  • 6.L.aMultilateral treaties and conventions
  • 6.L.bExport controls on frontier compute and models
  • 6.L.cCompute governance and licensing
  • 6.L.dSafety thresholds codified internationally

Design

Advancing15 signals
CHT recommends
  • Publish voluntary frontier-safety commitments that are specific and testable.
  • Share dangerous-capability evaluations with peers and regulators.
  • Align evaluation protocols internationally.
Indicators we track
  • 6.D.aVoluntary industry safety commitments
  • 6.D.bInformation sharing on dangerous capabilities
  • 6.D.cEvaluation protocols aligned internationally

Recent signals

AdvancingLaws · 6.L.aGLOBALMay 5, 2026

Commission services sign cooperation arrangement with Japan's Ministry of Internal Affairs and Communications to support the enforcement of digital platform regulation

The European Commission and Japan's Ministry of Internal Affairs and Communications have signed a cooperation arrangement to support the enforcement of digital platform regulations.

WhyEU and Japan sign cooperation arrangement to support enforcement of digital platform regulations, strengthening cross-border governance.Multilateral treaties and conventions
MixedMajorDesign · 6.D.aGLOBALApr 26, 2026

Our principles

OpenAI published a new set of guiding principles for its AI development, signaling a high-level commitment to safety and ethics.

WhyOpenAI (frontier lab) published new principles; represents a voluntary commitment but lacks specific, testable enforcement mechanisms.Voluntary industry safety commitments