2026-06-02
US order on frontier-model security
About this eventThe government established a voluntary secure early-access framework and classified evaluation of advanced models' cyber capabilities.
Directions · LAW AND AUDIT
AI regulation
In one viewGovernments started writing laws about what AI may and may not do and who answers for mistakes. It is both a brake on the industry and the condition for market access.
AI governance moved from declarations into operating mode: requirements, audits, reporting, fines. Deadlines and boundaries keep shifting.
Status
TypeDirections
Marker
Events in dossier7
Development chronology
Researched
2026-08-02
AI Act: rules take effect
About this eventThis deadline has passed: the main provisions and new transparency requirements became applicable on 2 August 2026. Specific high-risk-system duties follow their own transition schedule.
The AI Act enters into application in stages, so one date does not activate every obligation at once. The important shift is from voluntary principles to legal requirements for documentation, transparency, risk assessment and oversight; each system category must be checked against the Commission's current timeline.
Source: European CommissionIn progress
now
Chips as foreign policy
About this eventAccess to advanced accelerators remains an export-control instrument: the geography of compute is not set by the market.
Planned
2026-12-02
Labelling for pre-existing systems
About this eventThe limited grace period for machine marking and detection of synthetic content ends for systems released before 2 August 2026.
2027-12-02
Annex III rules
About this eventRequirements for stand-alone high-risk systems in employment, biometrics, justice and border control apply no later than this date.
planned
Capacity to evaluate
About this eventEurope plans its own pre-market model evaluation capacity, expected to be operational around 2027.
Distant horizons
ahead
An international regime
About this eventThe equivalent of non-proliferation treaties for systems above a capability threshold. Neither a body nor a verification mechanism exists.
Sources and research
Primary material behind this dossier: papers, lab publications and official reports.
Enabled by