Early AI Regulation
The GDPR (effective May 2018) became Europe's foundational data protection framework. In the U.S., concerns about algorithmic bias in hiring, lending, and criminal justice mounted. The FDA began reviewing AI/ML algorithms.
FoundationalStrategy Documents
The EU published the AI Act proposal (April 2021) with a risk-based tiered approach. The U.S. took a lighter approach. China published AI Ethics Guidelines. Self-regulation was the assumption.
StrategicChatGPT Launch
OpenAI released ChatGPT. 100 million users in 2 months. The regulatory landscape faced a general-purpose AI system accessible to anyone. Aggressive AI governance began.
CatalystGlobal Regulatory Scramble
UK AI Safety Summit at Bletchley Park. EU Parliament adopted AI Act position. China published GenAI regulations. Biden signed EO 14110. NIST published AI RMF 1.0.
Regulatory SurgeBiden EO + EU AI Act
EO 14110 directed agencies to establish AI safety standards. The EU AI Act published July 12, entered force August 1, 2024. Biden pursued AI chip export controls targeting China.
Federal ActionSB 1047 Veto
California SB 1047 passed both chambers but Newsom vetoed it. Several related AI bills did pass. National debate sparked: Should states or the federal government regulate AI?
Jurisdictional ClashInternational Divergence
EU: Strictest rules, 7% turnover penalties. China: "Core Socialist Values" alignment. UK: Light-touch approach. U.S.: Fragmented between federal and state.
Global FragmentationTrump EO Reversal
Trump signed EO 14179. Biden EO 14110 rescinded. AI Safety Institute gutted and rebranded. Federal agencies directed to reduce AI regulations.
DeregulationState vs. Federal War
Colorado SB 24-205 signed. New York expanded AI auditing (RAISE Act). Trump created DOJ AI Litigation Task Force to challenge state laws. Constitutional showdown looms.
Ongoing TensionEU GPAI Obligations In Force
General-purpose AI transparency obligations under the EU AI Act took effect August 2, 2025. GPAI providers must comply with transparency and copyright rules. Final Code of Practice published July 2025.
EU EnforcementTAKE IT DOWN Act
The first federal US law directly regulating AI-generated content entered enforcement. Platforms must remove nonconsensual intimate imagery and AI deepfakes within 48 hours. FTC enforcement began May 19, 2026.
Federal EnforcementColorado AI Law Replaced
SB 26-189 signed May 14, replacing SB 24-205 with a narrower ADMT notice-and-explanation framework effective January 2027. SB 24-205 reached its June 30 effective date without ever taking effect.
State LegislationWhite House AI Innovation and Security EO
Trump administration signed an executive order requiring voluntary pre-release government review of frontier AI models, the most significant US AI governance action since EO 14179. The order enabled the restriction of OpenAI's GPT-5.6 to trusted partners.
Federal ActionEU High-Risk Deadline Delayed
EU Digital Omnibus agreement pushed high-risk AI system obligations back 16 months -- from August 2, 2026 to December 2, 2027. The first formal change to the EU AI Act since adoption. GPAI obligations remain in effect.
EU Policy ShiftThe Current Landscape
US government actively restricts frontier model access. EU high-risk deadline pushed to December 2027. State laws advancing across the US. The era of unregulated AI is ending.
OngoingWhy This Matters
If you are building AI systems, regulatory clarity has become a business concern:
- Know your audience. A model compliant in America might not work within the EU.
- Plan for compliance costs. Testing, documentation, and auditing are now expected.
- Watch the U.S. federal-state fight. The outcome determines patchwork vs. unified rules.
- Understand your liability. Enforcement actions and lawsuits will follow.
- Engage with policy. The landscape is still forming. Industry input shapes outcomes.
Sources and Further Reading
- GDPR - Official Text
- EU AI Act - Full Text (Official Journal)
- EU AI Act Implementation Timeline
- EU Digital Omnibus - AI Act Amendment (High-Risk Delay to Dec 2, 2027)
- Biden Executive Order 14110 (October 2023)
- Trump Executive Order 14179 (January 2025)
- White House June 2026 EO on AI Innovation and Security
- California SB 1047 - Full Text
- Colorado SB 24-205 - Consumer Protections for AI
- Colorado SB 26-189 - ADMT Notice Framework (replaced SB 24-205)
- TAKE IT DOWN Act - Enforced since May 19, 2026
- Colorado SB 26-189 (ADMT notice framework, effective January 2027)
- NY RAISE Act (S6953-B) - Frontier AI developer requirements
- NIST AI Risk Management Framework (AI RMF 1.0)
- UK AI Regulation - Pro-Innovation Approach
- UK AI Security Institute (AISI) - Frontier AI Evaluations
This timeline is maintained as a public reference. Last updated: 2026-07-02.