Rules aimed at general-purpose models and the companies that build them.
A newer layer of rules targets the largest general-purpose and frontier models directly. The EU AI Act imposes transparency, documentation, and copyright duties on all general-purpose AI model providers from August 2025, with added systemic-risk obligations above a compute threshold. California's Transparency in Frontier Artificial Intelligence Act requires the largest developers to publish safety frameworks and report incidents. China requires security assessment and filing for public-facing generative services. South Korea adds duties for large-scale advanced AI.
If you build or fine-tune large models rather than just deploy them, this is the category written for you, and it sits on top of the deployment rules everyone else follows.
ChinaExtensive binding sectoral rules; no single comprehensive actIn force
European UnionComprehensive, horizontal, risk-based statutePhasing in
South KoreaComprehensive risk-based act with a light enforcement postureIn force
United KingdomSectoral, principles-based; no horizontal AI act by designNo comprehensive law
United States (Federal)No comprehensive law; deregulatory posture and state preemption pushNo comprehensive law
United States: CaliforniaSector-specific statutes; frontier transparency and content labelingPartially in force
JapanInnovation-first framework law; light-touch and non-punitiveIn force
United States: New YorkFrontier transparency law plus NYC hiring bias-audit rulePartially in force
VietnamComprehensive, risk-based AI law with extraterritorial reachIn force
SingaporeVoluntary frameworks and testing tools; no binding AI lawNo comprehensive lawThe Commission set out a coordinated plan to help member states and businesses handle the security risks of the most advanced AI models, and signaled a call to build EU capacity to evaluate models before they reach the market, expected operational by 2027.
Governor Pritzker signed the Artificial Intelligence Safety Measures Act (SB 315), putting Illinois beside California and New York with the same frontier template: models trained past 10^26 operations, developers above 500 million dollars in revenue, safety frameworks, transparency reports, and incident reporting. Illinois goes one step further than either coast: from January 1, 2028, large frontier developers must retain an independent third party to audit their compliance annually and publish the results. The act takes effect January 1, 2027, enforced exclusively by the attorney general, with no private right of action, and it preempts local AI regulation inside the state.
The AI Basic Act and its Enforcement Decree came into force. It sets transparency and high-impact duties, reaches foreign providers serving Korean users, and runs a one-year grace period before most fines apply. It joins the EU AI Act as one of only two comprehensive regimes in force.
California's frontier transparency and training-data laws, Texas TRAIGA, and Illinois' AI employment amendment all took effect on the same day, a month after the federal preemption order. For now every one of them remains enforceable, and most follow the affected resident rather than the company's location.
Artificial Intelligence Regulations, "Generative AI and foundation models in AI Regulation," reviewed July 28, 2026, https://artificialintelligenceregulations.com/topics/generative-foundation-models.html.
Entries state the position as of the review date and link to the primary text for verification.