AI Regulation: EU AI Act, US States, UK and Korea — 2026-09-13
As of September 2026, the United States remains without a comprehensive federal AI Act, forcing companies to navigate a complex patchwork of state laws in California, Texas, and Colorado. Meanwhile, the EU AI Act’s general applicability began on August 2, 2026, with the Digital Omnibus deferring specific high-risk obligations to late 2027 and 2028. In Korea, the AI Basic Act's implementation guidelines are being scrutinized for their ambiguity regarding "high-impact AI" classifications, while OpenAI has shifted its stance to support legally binding safety regulations.
AI Regulation: EU AI Act, US States, UK and Korea — 2026-09-13
Top developments
US State Patchwork Intensifies Without Federal Preemption
As of September 2026, the United States still lacks a comprehensive federal AI Act, leaving operators to face a live state patchwork. California SB 53 is already in force, Texas TRAIGA has been active since January, and Colorado’s rewritten framework is scheduled to start in January 2027. At least five states now have comprehensive AI laws either in force or effective soon, alongside over 40 narrower laws covering deepfakes, hiring, and chatbots.
EU Digital Omnibus Defers High-Risk Obligations
The EU AI Act became generally applicable on August 2, 2026, but the Digital Omnibus (Regulation (EU) 2026/1744) has modified its timeline. Specifically, Chapter III, Sections 1–3 regarding high-risk AI obligations have been deferred to December 2027 and August 2028. This adjustment provides additional time for compliance teams to address Annex III obligations, although transparency requirements under Article 50 remain live.
Korea Faces Ambiguity in "High-Impact" AI Classification
South Korean businesses are struggling to interpret the new AI Basic Act, which came into effect in January 2026. A recent expert analysis highlights the uncertainty surrounding the definition of "high-impact artificial intelligence" across eleven designated domains. Startups, particularly those developing recruitment screening tools, are facing difficulties determining whether their systems fall under these stringent regulatory categories, creating a compliance bottleneck.
OpenAI Shifts Stance on Federal Safety Regulations
OpenAI has officially called on the U.S. Congress to enact legally binding safety regulations for AI systems. This marks a significant policy shift for the company, which previously opposed certain regulatory measures. The change in stance follows recent controversies involving AI control lapses, signaling a potential alignment between leading AI developers and stricter federal oversight frameworks.
Local view
In South Korea, local media outlets like Seoul Finance are highlighting the practical challenges faced by startups under the new AI Basic Act. The discourse focuses on the lack of clarity in the enforcement decree regarding what constitutes "high-impact AI," with legal experts urging the Ministry of Science and ICT to provide more concrete examples for sectors like HR tech.
Context & numbers
- US State Laws: As of July 2026, at least 5 states (California, Colorado, Texas, New York, Illinois) have comprehensive AI laws in force or effective by January 2027.
- EU Enforcement: The European Commission’s AI Office began enforcing the AI Act alongside national authorities on August 2, 2026.
- Korean Penalties: Violations of AI content labeling obligations under the Korean AI Basic Act can result in corrective orders and fines of up to 30 million KRW.
On the radar
- Colorado Implementation: Companies should monitor Colorado’s rewritten AI framework, which is scheduled to go into effect in January 2027, as it replaces previous ADMT rules.
- EU High-Risk Deadlines: Compliance teams must prepare for the deferred high-risk AI obligations, which will become applicable in December 2027 and August 2028 following the Digital Omnibus amendments.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
