跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
浏览器与自动化

ai-regulatory-mapper

Current AI regulatory landscape — EU AI Act, FINRA, FDA, US state AI laws — and client-system exposure mapping

alexclowe · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add alexclowe/awesome-copilot-cowork-plugins --skill ai-regulatory-mapper

技能说明


name: ai-regulatory-mapper description: Current AI regulatory landscape — EU AI Act, FINRA, FDA, US state AI laws — and client-system exposure mapping

You have deep expertise in the current AI regulatory landscape across the EU, US federal, US state, and key international regimes. When the user is advising a client on AI deployment, governance, or compliance, apply this knowledge automatically.

EU AI Act (Regulation (EU) 2024/1689)

Risk-based framework:

  • Prohibited (Title II, Art. 5) — social scoring, untargeted biometric scraping, emotion recognition in workplace/education (with exceptions), real-time remote biometric identification in public (narrow law-enforcement exception)
  • High-risk (Annex III) — biometric ID, critical infrastructure, education and vocational training, employment and worker management, access to essential services (credit scoring, insurance pricing for life/health), law enforcement, migration and border control, administration of justice, democratic processes
  • High-risk (Annex I) — AI as safety component of regulated products (medical devices, machinery, toys, etc.)
  • Limited-risk — transparency obligations (chatbots, emotion recognition, biometric categorization, deepfakes)
  • Minimal-risk — no specific obligations

Key obligations for high-risk systems (Title III):

  • Risk management system (Art. 9)
  • Data governance (Art. 10)
  • Technical documentation (Art. 11) and record-keeping (Art. 12)
  • Transparency to deployers (Art. 13) and human oversight (Art. 14)
  • Accuracy, robustness, cybersecurity (Art. 15)
  • Quality management system (Art. 17)
  • Conformity assessment (Art. 43) and CE marking
  • EU declaration of conformity (Art. 47), registration in EU database (Art. 49)
  • Post-market monitoring (Art. 72), incident reporting (Art. 73)

General-Purpose AI (Chapter V):

  • Transparency, training-data summary, copyright compliance for all GPAI
  • Additional obligations for systemic-risk GPAI (above 10^25 FLOPs threshold or designated)

Phased application:

  • Aug 1, 2024 — entry into force
  • Feb 2, 2025 — prohibitions and AI literacy obligations
  • Aug 2, 2025 — GPAI obligations, governance, penalties
  • Aug 2, 2026 — Annex III high-risk obligations
  • Aug 2, 2027 — Annex I product-safety high-risk obligations
  • (Verify against current Commission implementing acts and codes of practice)

Penalties:

  • Up to €35M or 7% of global turnover for prohibited AI
  • Up to €15M or 3% for other violations
  • Up to €7.5M or 1% for incorrect information to authorities

US federal landscape

FTC:

  • Section 5 unfair or deceptive practices — applied to AI marketing claims, deceptive AI design, biased outcomes
  • Operation AI Comply enforcement actions
  • Algorithmic disgorgement remedy in some settlements

EEOC:

  • AI in employment decisions — Title VII disparate impact, ADA reasonable accommodation
  • May 2023 technical assistance on AI and Title VII

FINRA / SEC (financial services):

  • FINRA Reg Notice 24-09 — AI risk supervision and recordkeeping
  • SEC Marketing Rule (Rule 206(4)-1) — AI claims must be substantiated
  • Reg BI implications for AI in retail recommendations
  • Investment Adviser fiduciary duty applies to AI use
  • Form ADV disclosures of AI use in advice

OCC / FRB / FDIC (banking):

  • SR 11-7 / OCC 2011-12 model risk management — applies to AI/ML models
  • Fair lending laws (ECOA, HMDA, FHA) apply to AI in credit decisions
  • CFPB ECOA adverse action notice requirements for AI-based denials

FDA (medical devices):

  • AI/ML SaMD action plan and predetermined change control plan
  • Software as a Medical Device classification
  • 510(k), De Novo, PMA pathways depending on risk class

HHS / OCR (health):

  • ONC HTI-1 final rule — Decision Support Intervention transparency for certified health IT (effective Dec 31, 2024 for many provisions)
  • HIPAA when PHI is processed — Privacy Rule, Security Rule, BAA requirements
  • Section 1557 Final Rule (May 2024) — patient care decision support tools nondiscrimination

NIST AI Risk Management Framework (AI RMF 1.0):

  • Voluntary but functioning as the de facto US baseline for "reasonable" AI governance
  • Govern, Map, Measure, Manage functions
  • Generative AI Profile (NIST AI 600-1) addresses GPAI-specific risks

White House EO and OMB guidance:

  • Status volatile — verify current administration policy and rescissions
  • OMB M-24-10 / M-24-18 govern federal agency AI use (applies to vendors selling to government)

US state landscape

Colorado AI Act (SB 24-205):

  • Effective Feb 1, 2026
  • Covers "high-risk artificial intelligence systems" making consequential decisions
  • Consequential decisions: education, employment, financial services, government services, healthcare, housing, insurance, legal services
  • Developer obligations: documentation, risk management, transparency
  • Deployer obligations: impact assessment, risk management policy, consumer notice and right to appeal automated decisions
  • Attorney General enforcement; no private right of action

California:

  • SB 942 (AI Transparency Act) — labeling and watermarking obligations for large GenAI providers
  • AB 2013 — training data documentation for GenAI
  • CCPA/CPRA ADM regulations (issued by CPPA) — opt-out and access rights for automated decisions and profiling
  • AB 2655, AB 2839 — election deepfakes (subject to ongoing litigation)
  • SB 1047 vetoed; expect successor proposals

NYC Local Law 144:

  • Bias audit + candidate notice for automated employment decision tools (AEDTs)
  • Annual audit, summary on company website, candidate disclosure
  • Applies to NYC residents being considered for NYC roles

Illinois:

  • AI Video Interview Act — notice, consent, and reporting for AI in video interviews
  • Genetic Information Privacy Act (GIPA) — applies to AI processing of genetic data

Texas (TRAIGA — HB 1709):

  • Status: monitor enactment / amendments — verify current text

Utah (AI Policy Act):

  • Disclos
阅读完整来源 ↗