Data Overview

• Snapshot Week: 2026-05-29 to 2026-06-12
• Tracker: AI Regulation & Policy Tracker (view all snapshots: /policy/ai-regulation/data/?tracker=ai-regulation-tracker)
• Update Frequency: Weekly
• Primary Sources: EU AI Act Official Timeline, Fisher Phillips, UK Government, HK Law

Key Facts

• Who: US Congress (Great American AI Act), European Commission (Omnibus), UK AISI, GPAI signatories
• What: 20 regulatory entries across 5 jurisdictions; 4 Critical/High impact changes this week
• When: Week of June 12, 2026
• Impact: First comprehensive US federal AI framework proposal, largest EU compliance timeline extension since Act passage, Meta’s high-profile GPAI rejection

Week in Review

Four developments this week reshape the global AI regulatory landscape:

1. Great American AI Act of 2026 Emerges (June 4)

The first comprehensive federal AI framework proposal entered Congress. The 269-page discussion draft, introduced by Reps. Obernolte (R-CA) and Trahan (D-MA), proposes a 3-year preemption of state AI development laws — but notably not deployment. This distinction preserves the Colorado AI Act, California ADMT, NYC Local Law 144, and Illinois AI Video Interview Act. The bill also amends the WARN Act to require AI-specific layoff disclosures.

2. EU AI Act Omnibus Extends Deadlines (May 7 Agreement)

The largest timeline adjustment since the AI Act passed: Annex III high-risk systems deadline extended from August 2, 2026 to December 2, 2027 (16-month delay). Annex I product-embedded AI moved to August 2, 2028 (12-month delay). A new prohibition on AI-generated CSAM was added, with compliance deadline December 2, 2026.

3. UK AISI Rebrands to “AI Security Institute”

The name change from “AI Safety Institute” to “AI Security Institute” signals an intensified national security focus. The institute simultaneously formed a bilateral pact with Australia (May 25) for frontier AI capability sharing and evaluation collaboration.

4. GPAI Code of Practice Signatory Split

Approximately 24 organizations signed the GPAI Code of Practice, including Anthropic, Google, Microsoft, OpenAI, Amazon, IBM, Mistral AI, and Cohere. Meta declined entirely — no alternative compliance path announced. xAI signed only the Safety & Security chapter. Signatories receive presumption of conformity with EU AI Act obligations.

Trend Analysis

Trend 1: Federal-State Preemption Intensifies

US Executive Order 14365 (December 2025) directed the Attorney General to establish an AI Litigation Task Force to challenge “onerous” state AI laws, specifically targeting the Colorado AI Act. The Great American AI Act’s narrower 3-year preemption (development only) represents a legislative counterweight. The tension between federal standardization and state-level innovation protection will define US AI governance through 2027.

Trend 2: EU Compliance Timeline Extended

The 16-month Omnibus extension acknowledges a harmonized standards infrastructure gap. Only 8 of 27 Member States designated competent authorities by the August 2025 deadline. Germany (Bundesnetzagentur), Spain (AESIA), and Ireland (15 authorities) lead in framework development. Enforcement fragmentation will create an uneven landscape for at least 18 months.

Trend 3: International Coordination Accelerating

Three coordination mechanisms emerged this quarter:

• UK-Australia AI Security Pact (May 25 MOU)
• OpenAI and Microsoft joined UK’s International AI Coalition (February 2026)
• July 2026 AISI meeting expected to produce evaluation best practices — the first international standardization effort for frontier AI testing

Trend 4: Industry Alignment Diverges

GPAI Code of Practice signatures reveal a split: major labs (Anthropic, OpenAI, Google, Microsoft) committed to voluntary frameworks, while Meta took a high-profile rejection stance. xAI’s partial signature (Safety & Security chapter only) suggests selective compliance may become a pattern for newer entrants.

Trend 5: National Security Framing Dominates

UK’s AISI rebranding mirrors a broader shift: AI governance discourse is moving from “safety” (technical alignment, existential risk) toward “security” (national competitiveness, adversarial threats, critical infrastructure). Expect increased scrutiny of AI exports, talent flows, and infrastructure access.

This Week’s Data

European Union (6 Entries)

Date	Regulation/Policy	Type	Status	Impact	Key Details
2026-05-07	EU AI Act Omnibus — High-Risk Deadline Extension	Regulation	📋 Proposed	🔴 Critical	Annex III high-risk deadline extended from Aug 2, 2026 to Dec 2, 2027 (16-month delay); Annex I product-embedded AI moved to Aug 2, 2028 (12-month delay)
2026-08-02	EU AI Act — Full Enforcement Date	Regulation	⏳ Scheduled	🔴 Critical	Remaining provisions apply; operators of high-risk AI systems must comply; Member States must have operational regulatory sandboxes
2026-12-02	EU AI Act — AI-Generated CSAM Prohibition	Regulation	⏳ Scheduled	🟠 High	Digital Omnibus adds Article 5 prohibition: AI systems generating CSAM or non-consensual intimate imagery banned
2026-12-02	EU AI Act — AI-Generated Content Transparency	Guidelines	⏳ Scheduled	🟠 High	Transparency obligations for AI-generated content (watermarking, labeling); grace period reduced from 6 to 3 months
2026-06-09	GPAI Code of Practice Signatory Status	Framework	✅ In-Effect	🟠 High	~24 signatories; Meta declined; xAI partial (Safety & Security chapter only); signatories receive presumption of conformity
2026-06-09	EU Member State Authority Readiness Gap	Guidelines	✅ In-Effect	🟡 Medium	Only 8/27 Member States designated competent authorities by Aug 2025 deadline; 12 missed deadline entirely

United States — Federal (5 Entries)

Date	Regulation/Policy	Type	Status	Impact	Key Details
2026-06-04	Great American AI Act of 2026	Bill	📋 Proposed	🔴 Critical	First comprehensive federal AI framework (269 pages); 3-year state preemption for AI development (not deployment); WARN Act amendment for AI layoffs
2026-06-04	Great American AI Act — State Preemption Scope	Bill	📋 Proposed	🟠 High	Preemption applies to AI development only; Colorado AI Act, California ADMT, NYC Local Law 144, Illinois AI Video Interview Act remain intact
2026-06-04	Great American AI Act — WARN Act Amendment	Bill	📋 Proposed	🟠 High	When AI is a “substantial factor” in mass layoffs, WARN notice must: identify AI involvement, describe AI type/usage, estimate % of job losses attributable to AI, explain upskilling steps
2025-12-11	Executive Order 14365 — Federal AI Policy	Framework	✅ In-Effect	🔴 Critical	AG to establish AI Litigation Task Force (30 days) to challenge state laws conflicting with federal policy; Commerce to evaluate “onerous” state AI laws (90 days)
2025-12-11	Executive Order 14365 — Colorado AI Act Targeted	Framework	✅ In-Effect	🟠 High	EO specifically targets Colorado AI Act (effective Jun 30, 2026) as requiring AI models to “produce false results” through algorithmic discrimination obligations

United Kingdom (5 Entries)

Date	Regulation/Policy	Type	Status	Impact	Key Details
2026-01-29	AI Safety Institute Rebranding	Framework	✅ In-Effect	🟡 Medium	Rebranded from “AI Safety Institute” to “AI Security Institute”; signals national security pivot
2026-05-25	UK-Australia AI Security Pact	Framework	✅ In-Effect	🟠 High	MOU for frontier AI capability sharing, evaluation best practices, staff exchanges; signed in Canberra
2026-02-19	OpenAI/Microsoft Join UK AI Coalition	Framework	✅ In-Effect	🟠 High	Major labs join international coalition to safeguard AI development; builds on bilateral partnerships
2025-12-18	AISI Frontier AI Trends Report	Framework	📄 Published	🟠 High	First public evidence-based assessment of frontier AI evolution; 2 years of model testing; cybersecurity capabilities advancing rapidly
2026-01-28	AI Impact on UK Labour Market	Guidelines	📄 Published	🟡 Medium	Assessment of AI capabilities and workforce impact; research to inform policy decisions

International (2 Entries)

Date	Regulation/Policy	Type	Status	Impact	Key Details
2026-07	AISI Meeting — Evaluation Best Practices	Framework	⏳ Scheduled	🟠 High	July 2026 meeting expected to produce first international standardization effort for frontier AI testing and evaluation
2026-05-25	International Network for Advanced AI Measurement	Framework	✅ In-Effect	🟠 High	UK AISI shares best practices with AI research bodies across major economies; bilateral partnerships accelerating

China (1 Entry — Partial)

Date	Regulation/Policy	Type	Status	Impact	Key Details
2026-06-12	China AI Governance Updates	Guidelines	ℹ️ Informational	ℹ️ Informational	DATA INCOMPLETE: API rate limit prevented collection this week. CAC enforcement actions and algorithm registration updates not captured. Recommend manual follow-up.

Week-over-Week Summary

Metric	This Week	Last Week	Δ
Total entries	20	5	+15
Critical impact	4	1	+3
High impact	9	3	+6
Jurisdictions covered	5	3	+2
New entries	5	0	+5
Expanded entries	1	0	+1

Notable Entries Detail

Great American AI Act of 2026 — First Federal Framework

The 269-page discussion draft represents the most significant US legislative development this year. Key provisions:

Preemption Scope: The 3-year preemption applies only to AI development (building/training models), not deployment or use. This narrow scope preserves:

• Colorado AI Act (effective June 30, 2026)
• California ADMT regulations
• NYC Local Law 144
• Illinois AI Video Interview Act
• State privacy laws with AI provisions

WARN Act Amendment: When AI is a “substantial factor” in mass layoffs, employers must:

1. Identify AI involvement in the layoff notice
2. Describe the type and usage of AI
3. Estimate the percentage of job losses attributable to AI
4. Explain any upskilling or retraining steps taken

The Secretary of Labor has 300 days to issue implementing guidance.

Workforce Research Hub: Establishes an AI Workforce Research Hub to study AI’s impact on employment patterns.

EU AI Act Omnibus — Timeline Adjustments

The provisional agreement extends compliance deadlines for high-risk AI systems:

Annex III High-Risk Systems: August 2, 2026 → December 2, 2027 (16-month extension)

• Covers AI used in biometric identification, critical infrastructure, education, employment, law enforcement

Annex I Product-Embedded AI: August 2, 2027 → August 2, 2028 (12-month extension)

• Covers AI embedded in products covered by EU product safety legislation

New Prohibition: Digital Omnibus adds Article 5 prohibition on AI systems generating CSAM or non-consensual intimate imagery. Compliance deadline: December 2, 2026.

Transparency Grace Period: Reduced from 6 months to 3 months for AI-generated content labeling.

Member State Readiness Gap

Only 8 of 27 EU Member States designated competent authorities by the August 2025 deadline. The readiness gap creates enforcement fragmentation:

• Germany: Bundesnetzagentur leads with most advanced framework
• Spain: AESIA designated as national authority
• Ireland: 15 authorities designated across sectors
• 19 Member States: No designated authority or missed deadline

This fragmentation means enforcement intensity will vary significantly by jurisdiction until at least 2028.

GPAI Code of Practice — Industry Split

The General-Purpose AI Code of Practice signatory status reveals divergent compliance strategies:

Full Signatories (~24): Anthropic, Google, Microsoft, OpenAI, Amazon, IBM, Mistral AI, Cohere

• Receive presumption of conformity with EU AI Act GPAI obligations
• Access to harmonized standards development process

Declined: Meta

• No alternative compliance path announced
• High-profile rejection may signal strategic resistance

Partial Signatory: xAI

• Signed Safety & Security chapter only
• Selective compliance approach

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 65/100

The Great American AI Act’s development-only preemption carveout represents a calculated legislative compromise that most coverage treats as a technical detail. In reality, it creates a two-tier regulatory architecture: federal law governs model creation, while states regulate model application. This bifurcation has profound implications for the AI supply chain. Foundation model providers (OpenAI, Anthropic, Google) operate under a uniform federal regime, while deployers face a patchwork of state-level obligations. The WARN Act amendment’s 300-day guidance window is equally significant: it establishes AI as a legally cognizable cause of job displacement, creating new disclosure obligations that will ripple through corporate restructuring processes. The EU’s 16-month delay isn’t just a timeline extension — it’s an implicit admission that harmonized standards infrastructure doesn’t exist. With 19 Member States lacking competent authorities, the EU has effectively acknowledged enforcement will be uneven for at least 18 months, creating a compliance arbitrage window for sophisticated operators.

Key Implication: Companies building AI products for the US market should structure operations to separate development (federal protection) from deployment (state exposure), while EU-bound operations have an extended compliance runway but face fragmented enforcement until 2028.

Previous Snapshots

First snapshot for this tracker. Historical snapshots will appear here in future weeks.

Sources

• EU AI Act Official Timeline — EU AI Act Implementation, Updated 2026
• EU AI Act Phase 1 Implementation Update — Bits From Bytes, June 2026
• Great American AI Act Analysis — Fisher Phillips, June 4, 2026
• US Federal AI Regulation Analysis — Holland & Knight, December 2025
• UK AI Security Institute — UK Government, Updated 2026
• UK-Australia AI Security Pact — UK Government, May 25, 2026