AgentScout Logo Agent Scout

AI Regulation & Policy Tracker — Week of May 8, 2026

EU Omnibus trilogue stalled on high-risk AI compliance delays. US White House proposed federal AI preemption framework. Singapore launched first agentic AI governance framework. China enforcement actions ramping up for July deadline.

AgentScout · · · 8 min read
#ai-regulation #eu-ai-act #federal-ai-policy #agentic-ai #compliance
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

Data Overview

Key Facts

  • Who: EU institutions, US White House, China CAC, Singapore IMDA, Japan government
  • What: 17 active/pending regulations across 8 jurisdictions; 3 critical-impact developments this week
  • When: Snapshot covers May 2-8, 2026; key deadlines: EU GPAI enforcement (Aug 2, 2026), China platform compliance (July 1, 2026)
  • Impact: 4 critical-impact regulatory changes affecting AI deployers globally

Methodology

Data collected from official government sources, legislative tracking databases, and regulatory bodies. Each entry verified through primary sources where available (tier S/A), supplemented by legal analysis providers (tier B). Status classifications:

  • Proposed: Bill introduced or framework announced, not yet enacted
  • Announced: Official publication with implementation timeline
  • In-Effect: Legally binding and actively enforced
  • Delayed: Implementation postponed or negotiation ongoing

Impact levels assessed based on scope (jurisdiction reach), severity (compliance burden), and timeline urgency. Data current as of May 8, 2026.

This Week’s Data

Active and Pending Regulations

DateJurisdictionRegulation/PolicyTypeStatusImpactDetails
2026-04-29EUDigital AI Omnibus TrilogueRegulationDelayedCriticalEU countries and lawmakers failed to reach deal on postponing high-risk AI compliance deadlines; simplification measures still under negotiation
2026-04-28EUDigital AI Omnibus ProposalRegulationProposedCriticalSecond political trilogue on Omnibus to amend AI Act by postponing high-risk AI compliance deadlines
2026-03-31EUEnforcement of Chapter V (GPAI Models)GuidelinesAnnouncedHighDetailed guidance on Commission enforcement powers against GPAI model providers, effective August 2, 2026
2026-03-20US-FederalNational Policy Framework for AIFrameworkAnnouncedCriticalWhite House legislative framework urging Congress to establish federal preemption of state AI laws; six policy areas: child protection, electricity, IP, free speech, education
2026-03-20US-FederalH.R.5388 - American AI Leadership ActBillProposedHigh119th Congress bill for uniform federal AI regulation approach
2026US-FederalH.R.2152 - AI PLAN ActBillProposedMediumResources and technologies to combat AI use in financial crimes and misinformation
2026US-FederalH.R.1694 - AI Accountability ActBillProposedMediumAccountability requirements for AI systems in federal applications
2026US-StateState AI Legislation (45 states)BillProposedHigh635+ AI-related bills across 45 states in 2024-2026; carry-over from 2025 active; patchwork regulatory landscape
2026-07-01ChinaAI Compliance Framework for Digital PlatformsRegulationProposedCriticalNew compliance framework requiring product design review, content governance, user disclosure, youth protection; July 2026 implementation deadline
2026-02-01ChinaLivestreaming E-commerce SupervisionRegulationIn-EffectMediumSAMR/CAC measures requiring disclosure of AI-generated human representations in livestreaming
2025-09-01ChinaAIGC Labelling StandardsStandardIn-EffectHighMandatory labelling of AI-generated text, images, audio, video with visible and technical labels
2025-10-22UKInternational Scientific Report on AI SafetyFrameworkPublishedHighUK AISI interim report on advanced AI safety science, global handbook on AI safety
2026SingaporeModel AI Governance Framework for Agentic AIFrameworkAnnouncedHighIMDA launched new governance framework specifically for agentic AI systems
2025-09JapanAI Promotion ActAct/LawIn-EffectMediumJapan’s first AI legislation focusing on R&D promotion and utilization; no monetary penalties; National AI Committee chaired by President
2025-12-19JapanAI Utilization GuidelinesGuidelinesPublishedMediumGuidelines under AI Act Article 13; risk-based approach, stakeholder involvement, lifecycle governance framework
2026-08-02EUGPAI Model Provider Enforcement PowersRegulationAnnouncedCriticalCommission’s supervision and enforcement powers against GPAI model providers come into force; includes documentation requests, evaluations, fines
2027-08-02EUGPAI Legacy Model Compliance DeadlineRegulationAnnouncedCriticalGPAI models placed on market before Aug 2, 2025 must comply with AI Act obligations

Jurisdiction Summary

JurisdictionActive RegulationsPending BillsEnforcement ActiveNext DeadlineStatus
EU41No2026-08-02 (GPAI enforcement)Implementation Phase; Omnibus proposal stalled
US-Federal04NoUnknownFramework proposed; no comprehensive law passed
US-Statevaries635+ across 45 statesYesvariesPatchwork landscape; federal preemption debate
China31Yes2026-07-01 (Digital platforms)Enforcement ramping up; CAC active
UK00NoUnknownVoluntary frameworks; AISI evaluation platform active
Japan10NoUnknownAI Promotion Act operational; guidelines-based governance
Singapore10NoUnknownModel frameworks; agentic AI governance pioneer
International00NoUnknownUK AISI international report; bilateral agreements active

Week-over-Week Summary

MetricThis Week (May 8)Last Week (May 1)Change
Total entries tracked1716+1
Critical-impact items54+1
Jurisdictions covered880
In-effect regulations440
Proposed/pending items1312+1
New entries this week10+1

Key Changes This Week

JurisdictionRegulationChangeSignificance
EUDigital AI Omnibus TrilogueStatus changed to “Delayed (No Deal)“Major delay in proposed simplification of AI Act compliance; uncertainty for high-risk AI deployers
US-FederalNational Policy Framework for AIStatus changed from “Not Announced” to “Announced”First comprehensive federal AI legislative proposal with preemption focus; contrasts with 635+ state bills
SingaporeModel AI Governance Framework for Agentic AIStatus changed from “Not Announced” to “Announced”First dedicated governance framework for autonomous AI agents globally
ChinaAI Compliance FrameworkStatus advanced from “Draft” to “Proposed (July 2026 deadline)“Enforcement approaching for digital platforms; compliance preparation intensifying
  • EU Omnibus Stalled: The April 28-29 trilogue’s failure to reach agreement leaves high-risk AI compliance timelines uncertain. Companies cannot rely on postponement and must prepare for existing AI Act deadlines. The next trilogue date remains unannounced.

  • US Federal Preemption Intensifies: The White House National Policy Framework marks the first serious federal attempt to preempt state-level AI regulation. With 635+ state bills across 45 jurisdictions, the patchwork compliance burden has become a legislative priority. However, no comprehensive federal law has passed yet.

  • China Enforcement Accelerating: Shanghai and Zhejiang CAC have already begun penalizing unregistered AI services. The July 1, 2026 deadline for digital platform compliance creates a 54-day countdown for affected companies.

  • Singapore Pioneers Agentic AI Governance: The Model AI Governance Framework for Agentic AI is the first jurisdiction-specific regulation addressing autonomous AI agents. This positions Singapore as a regulatory leader for the next generation of AI systems.

  • Japan’s Innovation-First Approach: The AI Promotion Act (operational since September 2025) imposes no monetary penalties, focusing instead on R&D incentives and guidelines. This contrasts sharply with the EU’s enforcement-heavy model.

  • GPAI Compliance Clock: EU enforcement powers for general-purpose AI model providers activate August 2, 2026 (85 days from this snapshot). Companies placing GPAI models before August 2025 must comply by August 2027.

  • International Coordination: The UK AISI International Scientific Report on AI Safety (October 2025) provides a shared framework for AI safety evaluation. UK-US-Singapore bilateral partnerships remain active in AI safety coordination.

Notable Developments

  1. EU Omnibus Trilogue Failure (April 29, 2026): The second political trilogue on the Digital AI Omnibus failed to reach agreement on postponing high-risk AI obligations. This creates regulatory uncertainty for companies banking on extended compliance timelines. Critical implications for EU AI Act implementation.

  2. US White House National Policy Framework (March 20, 2026): The first comprehensive federal AI legislative framework was announced, urging Congress to establish federal preemption of state AI laws across six policy areas. This represents a significant shift toward centralized AI governance in the US.

  3. Singapore Agentic AI Framework: IMDA launched the world’s first dedicated governance framework for autonomous AI agents. This regulatory innovation addresses the unique risks of self-directed AI systems before other jurisdictions have begun grappling with the issue.

Key Deadlines Timeline

DeadlineJurisdictionRegulationAction Required
2026-07-01ChinaAI Compliance Framework for Digital PlatformsFull compliance required for platform operators
2026-08-02EUGPAI Model Provider Enforcement PowersDocumentation and evaluation readiness for GPAI providers
2027-08-02EUGPAI Legacy Model ComplianceGPAI models placed on market before Aug 2025 must comply

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 65/100

The EU Omnibus trilogue failure receives coverage as a negotiation update, but the operational implication is underreported: companies that paused compliance preparation in anticipation of extended timelines now face a 73-day gap before GPAI enforcement activates (August 2, 2026). For high-risk AI deployers, the compliance window has effectively closed for “wait and see” strategies. Meanwhile, the US federal preemption framework’s six policy areas (child protection, electricity, IP, free speech, education) map directly to the 312 state bills targeting those same domains—creating a preemption battleground that will intensify through the 2026 election cycle. Singapore’s agentic AI framework, while covered as a governance initiative, positions it as the regulatory sandbox for autonomous agent compliance—a first-mover advantage that will influence EU and US rulemaking on multi-agent systems.

Key Implication: Companies with multi-jurisdictional AI deployments should prioritize China platform compliance (54 days) and EU GPAI documentation (85 days) over waiting for regulatory clarity—the stasis in Brussels and the preemption debate in Washington will not resolve before these deadlines.

Previous Snapshots

Sources

AI Regulation & Policy Tracker — Week of May 8, 2026

EU Omnibus trilogue stalled on high-risk AI compliance delays. US White House proposed federal AI preemption framework. Singapore launched first agentic AI governance framework. China enforcement actions ramping up for July deadline.

AgentScout · · · 8 min read
#ai-regulation #eu-ai-act #federal-ai-policy #agentic-ai #compliance
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

Data Overview

Key Facts

  • Who: EU institutions, US White House, China CAC, Singapore IMDA, Japan government
  • What: 17 active/pending regulations across 8 jurisdictions; 3 critical-impact developments this week
  • When: Snapshot covers May 2-8, 2026; key deadlines: EU GPAI enforcement (Aug 2, 2026), China platform compliance (July 1, 2026)
  • Impact: 4 critical-impact regulatory changes affecting AI deployers globally

Methodology

Data collected from official government sources, legislative tracking databases, and regulatory bodies. Each entry verified through primary sources where available (tier S/A), supplemented by legal analysis providers (tier B). Status classifications:

  • Proposed: Bill introduced or framework announced, not yet enacted
  • Announced: Official publication with implementation timeline
  • In-Effect: Legally binding and actively enforced
  • Delayed: Implementation postponed or negotiation ongoing

Impact levels assessed based on scope (jurisdiction reach), severity (compliance burden), and timeline urgency. Data current as of May 8, 2026.

This Week’s Data

Active and Pending Regulations

DateJurisdictionRegulation/PolicyTypeStatusImpactDetails
2026-04-29EUDigital AI Omnibus TrilogueRegulationDelayedCriticalEU countries and lawmakers failed to reach deal on postponing high-risk AI compliance deadlines; simplification measures still under negotiation
2026-04-28EUDigital AI Omnibus ProposalRegulationProposedCriticalSecond political trilogue on Omnibus to amend AI Act by postponing high-risk AI compliance deadlines
2026-03-31EUEnforcement of Chapter V (GPAI Models)GuidelinesAnnouncedHighDetailed guidance on Commission enforcement powers against GPAI model providers, effective August 2, 2026
2026-03-20US-FederalNational Policy Framework for AIFrameworkAnnouncedCriticalWhite House legislative framework urging Congress to establish federal preemption of state AI laws; six policy areas: child protection, electricity, IP, free speech, education
2026-03-20US-FederalH.R.5388 - American AI Leadership ActBillProposedHigh119th Congress bill for uniform federal AI regulation approach
2026US-FederalH.R.2152 - AI PLAN ActBillProposedMediumResources and technologies to combat AI use in financial crimes and misinformation
2026US-FederalH.R.1694 - AI Accountability ActBillProposedMediumAccountability requirements for AI systems in federal applications
2026US-StateState AI Legislation (45 states)BillProposedHigh635+ AI-related bills across 45 states in 2024-2026; carry-over from 2025 active; patchwork regulatory landscape
2026-07-01ChinaAI Compliance Framework for Digital PlatformsRegulationProposedCriticalNew compliance framework requiring product design review, content governance, user disclosure, youth protection; July 2026 implementation deadline
2026-02-01ChinaLivestreaming E-commerce SupervisionRegulationIn-EffectMediumSAMR/CAC measures requiring disclosure of AI-generated human representations in livestreaming
2025-09-01ChinaAIGC Labelling StandardsStandardIn-EffectHighMandatory labelling of AI-generated text, images, audio, video with visible and technical labels
2025-10-22UKInternational Scientific Report on AI SafetyFrameworkPublishedHighUK AISI interim report on advanced AI safety science, global handbook on AI safety
2026SingaporeModel AI Governance Framework for Agentic AIFrameworkAnnouncedHighIMDA launched new governance framework specifically for agentic AI systems
2025-09JapanAI Promotion ActAct/LawIn-EffectMediumJapan’s first AI legislation focusing on R&D promotion and utilization; no monetary penalties; National AI Committee chaired by President
2025-12-19JapanAI Utilization GuidelinesGuidelinesPublishedMediumGuidelines under AI Act Article 13; risk-based approach, stakeholder involvement, lifecycle governance framework
2026-08-02EUGPAI Model Provider Enforcement PowersRegulationAnnouncedCriticalCommission’s supervision and enforcement powers against GPAI model providers come into force; includes documentation requests, evaluations, fines
2027-08-02EUGPAI Legacy Model Compliance DeadlineRegulationAnnouncedCriticalGPAI models placed on market before Aug 2, 2025 must comply with AI Act obligations

Jurisdiction Summary

JurisdictionActive RegulationsPending BillsEnforcement ActiveNext DeadlineStatus
EU41No2026-08-02 (GPAI enforcement)Implementation Phase; Omnibus proposal stalled
US-Federal04NoUnknownFramework proposed; no comprehensive law passed
US-Statevaries635+ across 45 statesYesvariesPatchwork landscape; federal preemption debate
China31Yes2026-07-01 (Digital platforms)Enforcement ramping up; CAC active
UK00NoUnknownVoluntary frameworks; AISI evaluation platform active
Japan10NoUnknownAI Promotion Act operational; guidelines-based governance
Singapore10NoUnknownModel frameworks; agentic AI governance pioneer
International00NoUnknownUK AISI international report; bilateral agreements active

Week-over-Week Summary

MetricThis Week (May 8)Last Week (May 1)Change
Total entries tracked1716+1
Critical-impact items54+1
Jurisdictions covered880
In-effect regulations440
Proposed/pending items1312+1
New entries this week10+1

Key Changes This Week

JurisdictionRegulationChangeSignificance
EUDigital AI Omnibus TrilogueStatus changed to “Delayed (No Deal)“Major delay in proposed simplification of AI Act compliance; uncertainty for high-risk AI deployers
US-FederalNational Policy Framework for AIStatus changed from “Not Announced” to “Announced”First comprehensive federal AI legislative proposal with preemption focus; contrasts with 635+ state bills
SingaporeModel AI Governance Framework for Agentic AIStatus changed from “Not Announced” to “Announced”First dedicated governance framework for autonomous AI agents globally
ChinaAI Compliance FrameworkStatus advanced from “Draft” to “Proposed (July 2026 deadline)“Enforcement approaching for digital platforms; compliance preparation intensifying
  • EU Omnibus Stalled: The April 28-29 trilogue’s failure to reach agreement leaves high-risk AI compliance timelines uncertain. Companies cannot rely on postponement and must prepare for existing AI Act deadlines. The next trilogue date remains unannounced.

  • US Federal Preemption Intensifies: The White House National Policy Framework marks the first serious federal attempt to preempt state-level AI regulation. With 635+ state bills across 45 jurisdictions, the patchwork compliance burden has become a legislative priority. However, no comprehensive federal law has passed yet.

  • China Enforcement Accelerating: Shanghai and Zhejiang CAC have already begun penalizing unregistered AI services. The July 1, 2026 deadline for digital platform compliance creates a 54-day countdown for affected companies.

  • Singapore Pioneers Agentic AI Governance: The Model AI Governance Framework for Agentic AI is the first jurisdiction-specific regulation addressing autonomous AI agents. This positions Singapore as a regulatory leader for the next generation of AI systems.

  • Japan’s Innovation-First Approach: The AI Promotion Act (operational since September 2025) imposes no monetary penalties, focusing instead on R&D incentives and guidelines. This contrasts sharply with the EU’s enforcement-heavy model.

  • GPAI Compliance Clock: EU enforcement powers for general-purpose AI model providers activate August 2, 2026 (85 days from this snapshot). Companies placing GPAI models before August 2025 must comply by August 2027.

  • International Coordination: The UK AISI International Scientific Report on AI Safety (October 2025) provides a shared framework for AI safety evaluation. UK-US-Singapore bilateral partnerships remain active in AI safety coordination.

Notable Developments

  1. EU Omnibus Trilogue Failure (April 29, 2026): The second political trilogue on the Digital AI Omnibus failed to reach agreement on postponing high-risk AI obligations. This creates regulatory uncertainty for companies banking on extended compliance timelines. Critical implications for EU AI Act implementation.

  2. US White House National Policy Framework (March 20, 2026): The first comprehensive federal AI legislative framework was announced, urging Congress to establish federal preemption of state AI laws across six policy areas. This represents a significant shift toward centralized AI governance in the US.

  3. Singapore Agentic AI Framework: IMDA launched the world’s first dedicated governance framework for autonomous AI agents. This regulatory innovation addresses the unique risks of self-directed AI systems before other jurisdictions have begun grappling with the issue.

Key Deadlines Timeline

DeadlineJurisdictionRegulationAction Required
2026-07-01ChinaAI Compliance Framework for Digital PlatformsFull compliance required for platform operators
2026-08-02EUGPAI Model Provider Enforcement PowersDocumentation and evaluation readiness for GPAI providers
2027-08-02EUGPAI Legacy Model ComplianceGPAI models placed on market before Aug 2025 must comply

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 65/100

The EU Omnibus trilogue failure receives coverage as a negotiation update, but the operational implication is underreported: companies that paused compliance preparation in anticipation of extended timelines now face a 73-day gap before GPAI enforcement activates (August 2, 2026). For high-risk AI deployers, the compliance window has effectively closed for “wait and see” strategies. Meanwhile, the US federal preemption framework’s six policy areas (child protection, electricity, IP, free speech, education) map directly to the 312 state bills targeting those same domains—creating a preemption battleground that will intensify through the 2026 election cycle. Singapore’s agentic AI framework, while covered as a governance initiative, positions it as the regulatory sandbox for autonomous agent compliance—a first-mover advantage that will influence EU and US rulemaking on multi-agent systems.

Key Implication: Companies with multi-jurisdictional AI deployments should prioritize China platform compliance (54 days) and EU GPAI documentation (85 days) over waiting for regulatory clarity—the stasis in Brussels and the preemption debate in Washington will not resolve before these deadlines.

Previous Snapshots

Sources

xs1el7m2v2t3yavlr83nq████g47qbo7gm7go4fv3ziqmyf8uv22l4vz09████tjk2q99jlnc7i2uh4kxq7xco8a15qabui████w37kadvtvifduv3gwb1b616419ng49z6░░░ejamdnu64qbzjttahsov0o3iiimriak13████af6oyut7jmpakw1ehjvscm4wy6pzivm░░░1oleupxl9bdi8z70mggjbpacte5x2lmdaa░░░611yknpmmp8kbb7x5ipbk8v179q49pamk░░░ckkna636p9sanwfzhacz8ou4f9vxypy2i████4sxggbpmqyeug9ex7cfw1xifs3nlfpka████cyobecfolx6yl4ro6733i87y5ifim4qr████7go9plydlmnusthecynsvecape901nfp░░░hqaatwcfg7t41478ac2ik56vfpedmaf0u████9qjkxd1vt68w4s7e8ungcil2u1g871pzs████01i0moguet7m3arh4tzsc4xx5bd68fpup████15167fn9uijzcou5qugs16w0ga2jty44░░░wt6ev5u28zm1l4dn8aqvnw603ifku0caw████05p7fqf3tfyejbdg9ysl7oblrf0whto44░░░omk2gwvwp0ee57z4h9q7inbpqlscb8gk░░░ggbxwbpbqr86g1vmmjg4fnv1mlrs1buae████n5sg3led4zzckhgf83zas4s3n2affi8████shrjtlnnxhfijuxj8x9jjncanhgz8j░░░nqptl5q868na0sdt63eljulbklkghres░░░s0h1zqm5x797tnfkx70m65ecj0cixbti░░░9t1huxwxjxc69wht5ic6locr1x1uqhtyn░░░w1npxd97depqvrfti0xo6clq56u5hv2ep████0slptm2h65egu0fxyq7716psb0v5mksz8░░░8yengjnxuitnjn7jtipmrei4ww01m2jek░░░odwnjy4gsa8mzmk9sn3oceqj1bfdgmxs░░░zp7ce97647nhpi14u3vuqh956fh4dlk░░░x9f603tmfyjvge6h4x3ldaah412fa04░░░2j8d6q13vrrkvs3am9ektyd6l44j3fn████7e6mpdxvxkn70qo02dyk97e225egz9jxl████y07xp23qdzu0xi5i5vu4urpmkmbuxq████0a1v5umswskanlrx659h9gfnzrvevlfoc████sngkvlz0fq1msun1wvz3w46ocf5qoa62████56eg95ejs8fskfz5gipfui40gijset751████k0b04p8h6ulylmrjn4cnmhxgcvui8up████lkrfiev8bhe9wlspn5jlyss9zk1e6meq████jxj4jgc4vzqbu8inthedg8s0x4jsbl9████plxcv1se011j7jjt2wlni3xv098osf░░░vlrryifn8rocyje4sblpesmr5rxqlb5░░░wtjcs0y1ekotwv9z2955ms64krua3hkls████5qa6nnhnodmxbr77119tlpu19mxlzir0i░░░qt5rurdnbjlb64gh6et8uu0yyjdw5m7qd████853rd20z7gpkdwpkxlfa4si2igpeblnp████h1p9xnqkvp8iq0qpi12p7wc6wyay80gb████r6k9rceja3forgxzlpv0le8p98ct3yq03████0ee8e3l9logsnnfczkarf09b7xkthfiw8s████8e182eodci34mykd1rcvqk11ywq26dlz1████9edvl76oqab