Pyyan / Compare / EU AI Office vs UK AI Security Institute vs CAISI vs CAC

EU AI Office vs UK AI Security Institute vs CAISI vs CAC

4 of 5

Regulators · verified 13 Aug 2026

×EEU AI OfficeEuropean Commissioncurrent
×UUK AI Security InstituteUnited Kingdomcurrent
×CAISIUnited Statescurrent
×CACChinacurrent
1 slot left
SpecificationEU AI OfficeUK AI Security InstituteCAISICAC
SummaryFirst hard enforcement regime for AI.Frontier model pre-deployment testing.Center for AI Standards and Innovation.Cyberspace Administration of China.
JurisdictionEuropean UnionUnited KingdomUnited StatesChina
InstrumentEU AI ActSectoralNIST AI RMFCybersecurity Law
Enforcing since2 Aug 2026202420241 Jan 2026
CategoryRegulatorsRegulatorsRegulatorsRegulators
OfficialEuropean CommissionUnited KingdomUnited StatesChina

Highlighted rows are where these differ.

EU AI Office

  • Enforcement began 2 August 2026
  • High-risk obligations live the same day
  • EU evaluation capacity expected 2027

Best for .

Full spec sheet →

UK AI Security Institute

  • Outcome-based, pro-innovation model
  • No single UK AI act

Full spec sheet →

CAISI

  • Renamed from the US AI Safety Institute
  • No federal AI law — state patchwork

Full spec sheet →

CAC

  • AI security review and data localisation
  • Fines and suspensions issued without published reasoning

Full spec sheet →