Data Overview

• Snapshot Week: 2026-06-12 to 2026-06-19
• Tracker: AI Regulation Tracker (view all historical snapshots: /policy/ai-regulation/data/?tracker=ai-regulation-tracker)
• Update Frequency: Weekly
• Primary Sources: EU AI Act Official, NIST CAISI, US Federal AI Bills, UK AI Security Institute, China AI Regulation Analysis

Key Facts

• Who: EU, US Federal Government, UK, Australia, China regulatory bodies
• What: 13 regulatory frameworks and policies tracked; 5 critical/high-impact developments this week; 850+ algorithms registered under China’s pre-approval regime
• When: Snapshot period Jun 12–19, 2026; key events spanning Jan 2023–Jun 2026
• Impact: 4 “Critical” impact regulations across EU, US, and China; first bilateral AI security MOU signed; first bipartisan US federal AI framework draft released

Methodology

Data collected via Jina Reader API with rate-limited Brave Search fallback. Each entry verified against primary source documentation. Regulatory status tracked through official government announcements and verified secondary sources. Impact levels assessed based on:

• Critical: Immediate compliance requirements affecting multi-national deployments, new regulatory bodies or frameworks, or binding enforcement mechanisms
• High: Significant policy shifts, international coordination mechanisms, or new guidance from regulatory bodies
• Medium: Technical guidance, research agreements, or non-binding frameworks

Data accurate as of 2026-06-19. Regulatory landscapes evolve rapidly; verify current status before compliance decisions.

This Week’s Data

Date	Jurisdiction	Regulation/Policy	Type	Status	Impact Level	Key Details
2026-05-07	EU	EU AI Act Omnibus - Political Agreement	Regulation	Agreed	Critical	High-risk systems timeline extended: biometrics/infrastructure Dec 2027, products Aug 2028; new ban on AI nudification apps; AI Office powers strengthened; SME/SMC simplifications extended
2026-06-04	US-Federal	Great American Artificial Intelligence Act (Discussion Draft)	Bill	Proposed	Critical	Bipartisan draft by Obernolte (R-CA) and Trahan (D-MA); nationalizes frontier model governance; CAISI authorization $100M/year; aims to preempt state-level AI patchwork
2026-05-25	UK	UK-Australia AI Security Pact (MOU)	Framework	In-Effect	High	UK AISI and Australian AISI to share frontier AI capability info, collaborate on evaluations, staff exchanges; focus on cybersecurity and national security risks
2026-01-29	UK	UK AISI Rebranding to ‘AI Security Institute’	Framework	In-Effect	High	Name change signals pivot to national security framing; mission redefined as ‘minimizing surprise from rapid AI advances’; cyber-attack capability focus
2026-05-01	US-Federal	NIST CAISI DeepSeek V4 Pro Evaluation	Guidelines	Published	High	CAISI evaluated open-weight DeepSeek V4; comparative assessment of US vs PRC model capabilities; benchmarks across five domains
2026-03-27	US-Federal	CAISI-OpenMined CRADA for Secure AI Evaluations	Framework	In-Effect	Medium	Collaborative research agreement to enable privacy-preserving AI evaluations; secure testing protocols development
2026-02-19	UK	OpenAI and Microsoft Join UK AI Coalition	Framework	In-Effect	High	OpenAI and Microsoft joined UK’s international coalition to safeguard AI development; voluntary framework participation
2023-08-15	China	Generative AI Measures (CAC)	Regulation	In-Effect	Critical	Mandatory pre-launch security assessment (3-6 months); algorithm registration; content filtering aligned with socialist values; real-name verification required; fines up to 10% revenue
2023-01-10	China	Deep Synthesis (Deepfake) Regulations	Regulation	In-Effect	High	Permanent labeling of synthesized content; identity verification for persons being synthesized; 24-hour takedown response; prohibited uses include impersonation without consent
2022-03-01	China	Algorithm Recommendation Registration System	Regulation	In-Effect	Critical	850+ algorithms registered as of Dec 2023; 2-4 month registration timeline; security assessment for >100M users; CAC can order algorithm modifications at any time; annual updates required
2025-12-11	US-Federal	Executive Order 14365 - Federal AI Policy Framework	Framework	In-Effect	Critical	Established federal AI governance baseline; agency coordination requirements; AI procurement standards; workforce protections
2024-08-01	EU	EU AI Act Entry into Force	Act/Law	In-Effect	Critical	Full applicability Aug 2026; prohibited practices effective Feb 2025; GPAI rules effective Aug 2025; 4-tier risk classification
2026-03-09	US-Federal	NIST AI 800-4: Monitoring Deployed AI Systems	Guidelines	Published	Medium	Report on challenges to monitoring deployed AI; practitioner workshop findings; systematic literature review; post-deployment surveillance guidance

Week-over-Week Summary

Metric	This Week	Last Week	Change
Total entries	13	5	+8
Critical impact	4	2	+2
High impact	4	2	+2
Medium impact	2	1	+1
Jurisdictions covered	4	3	+1 (Australia via UK pact)
New entries this week	2	0	+2

Trends & Observations

Global governance paradigms diverging rapidly. Four distinct regulatory philosophies are crystallizing: the EU’s risk-based framework with timeline simplifications for compliance practicality; the US bipartisan move toward federalized frontier model oversight with state preemption; the UK’s pivot from “safety” to “security” framing signaling national security prioritization; and China’s ideology-driven pre-approval regime with 850+ registered algorithms. This divergence creates compliance complexity for multi-jurisdictional deployments but also establishes natural experiments for regulatory effectiveness.

US federalization accelerates with bipartisan momentum. The Great American AI Act represents the first comprehensive attempt to create a federal AI governance framework with bipartisan sponsorship. The $100M/year CAISI authorization and explicit preemption language targeting state-level “patchwork” indicates congressional recognition that fragmented state regulations impede AI development. If enacted, this would create a unified US approach similar to the GDPR’s effect in harmonizing EU privacy rules.

UK repositions from safety to security framing. The rebranding of the UK AI Safety Institute to “AI Security Institute” in January 2026 marks a strategic shift toward national security prioritization. This reframing aligns with the UK-Australia bilateral MOU focused on “frontier AI capability sharing” for cybersecurity purposes, diverging from the EU’s rights-based approach. The UK is building an alternative regulatory coalition network centered on security cooperation rather than harmonized safety standards.

China’s pre-approval regime documented as world’s most comprehensive. With 850+ algorithms registered and mandatory 3-6 month security assessments before deployment, China’s algorithm registration system represents the most extensive pre-approval framework globally. The CAC’s authority to mandate algorithm modifications “at any time” creates ongoing compliance obligations absent in Western frameworks. This regime provides a reference point for understanding maximum-regulation governance approaches.

International coordination mechanisms multiplying. Beyond bilateral MOUs (UK-Australia), voluntary coalitions (UK AI Coalition with OpenAI/Microsoft participation), and evaluation frameworks (NIST CAISI’s DeepSeek assessments), new coordination channels are emerging. These mechanisms operate below formal treaty level but create de facto governance networks. The proliferation suggests recognition that unilateral national approaches cannot address transnational AI governance challenges.

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 60/100

The four regulatory paradigms—EU timeline extensions, US federalization, UK security pivot, China pre-approval—are not merely diverging but actively competing for global influence. The EU AI Act Omnibus timeline extension to Dec 2027/Aug 2028 represents the first major regulatory retreat from initially aggressive compliance timelines, acknowledging industry feedback on implementation feasibility. The Great American AI Act’s explicit preemption language targeting state-level AI laws signals a deliberate strategy to prevent California-style regulatory fragmentation from emerging in AI governance. The UK-Australia bilateral MOU marks the first documented frontier AI capability-sharing agreement between national AI institutes, creating a nascent security-focused regulatory bloc distinct from EU rights-based approaches. China’s 850+ registered algorithms represent the world’s only mandatory pre-deployment approval regime with documented compliance volumes—the regime is functional and scaling, not theoretical. NIST CAISI’s evaluation of DeepSeek V4 Pro indicates US agencies are actively assessing adversary model capabilities for security risks, suggesting national security considerations are embedded in technical evaluation frameworks even without explicit legislative mandates.

Key Implication: Multi-national AI deployments will face not just regulatory divergence but competing governance paradigms—compliance strategies designed for one paradigm may conflict with requirements in others, making jurisdiction-specific compliance architecture a competitive necessity rather than a regulatory afterthought.

Sources

• EU AI Act Official — European Commission
• NIST CAISI — National Institute of Standards and Technology
• US Federal AI Bills — House Committee on Science, Space, and Technology
• UK AI Security Institute — UK Government
• China AI Regulation Analysis — Pertama Partners, 2026