EU
Comprehensive
GPAI obligations incl. Annex XI energy/compute docs
Binds Providers of general-purpose AI models. Technical documentation, training summary, energy disclosure.
Stated maximum penalty — Up to 3% turnover or €15M
Topic dossier
Frontier & general-purpose AI obligations
The rules aimed at foundation-model and frontier developers — safety frameworks, incident reporting and model documentation. 7 obligations across 4 jurisdictions — 4 in force, 1 proposed. Next dated deadline: 2 Aug 2026.
A distinct regulatory track targets the largest models rather than specific uses. It splits into model-documentation duties for general-purpose AI (GPAI) and safety-and-incident regimes for frontier developers above a capability or compute threshold. The EU AI Act covers GPAI; California’s SB 53 and New York’s RAISE Act lead on frontier safety in the US. The instruments below make up this track.
Other themes DeepfakesTransparencyHigh-risk AIProhibitedHiring
The Register
7 obligations · 4 jurisdictionsEuropean Union 2
EU
Comprehensive
GPAI enforcement & fines (Art. 101)
Binds GPAI model providers (enforced by the Commission / AI Office). Commission may fine GPAI providers; supervision powers begin.
Stated maximum penalty — Up to 3% global turnover or €15M
United States 3
US · CA
Binding
SB 53 — Frontier AI Transparency Act
Binds Frontier AI developers (>1e26 training ops); large frontier developers (>$500M revenue). Safety frameworks + critical-incident reporting to Cal OES for frontier developers.
Stated maximum penalty — Up to $1M per violation
US · CA
Binding
AB 2013 — GenAI training-data transparency
Binds Developers of generative AI systems made available to Californians. Public dataset-summary disclosure for generative AI offered to Californians.
Stated maximum penalty — Civil enforcement
US · NY
Binding
RAISE Act — frontier AI safety
Binds Large frontier developers (>1e26 ops, >$500M revenue) operating in New York. Safety/security protocols + 72h incident reporting to NY AG/DHSES.
Stated maximum penalty — Up to $1M / $3M
China 1
China
Binding
Generative AI Interim Measures
Binds Providers of public-facing generative AI services in mainland China. Security assessment, algorithm filing and content controls for public-facing GenAI services.
Stated maximum penalty — Rectification, suspension, criminal referral
Canada 1
Canada
Proposed
Federal AIDA (Bill C-27) — abandoned
Binds Would have bound high-impact & general-purpose AI — never enacted. Would have set high-impact / GPAI duties; died on prorogation Jan 2025 and was not reintroduced.
Abandoned — died on prorogation Jan 2025; not reintroduced.
Questions & answers
From the dataWhat is GPAI under the EU AI Act?
General-purpose AI — a model that can perform a wide range of tasks and be integrated into many downstream systems. Providers must keep technical documentation, publish a training-content summary and, for models posing systemic risk, meet additional safety and reporting duties.
What do frontier-AI safety laws require?
Published safety frameworks and mandatory reporting of critical incidents. California’s SB 53 requires frontier developers to report incidents to Cal OES; New York’s RAISE Act requires security protocols and 72-hour incident reporting.
Is there a US federal frontier-AI law?
No comprehensive one. Frontier-specific obligations are currently state-led — California’s SB 53 and New York’s RAISE Act — while federal action remains limited to executive measures and sector rules.
Which jurisdictions does AI Law Radar track for frontier & general-purpose ai?
We currently track frontier & general-purpose ai obligations across 4 jurisdictions: European Union, United States, China and Canada. Each is dated and linked to its primary source on this page.
Not legal advice. Each obligation links to its primary source and carries the date it was last checked; verify the legal text before relying on it.