---
id: ai-governance-and-policy
title: "AI Governance: Risk Frameworks, Audits, and International Cooperation"
schema_type: TechArticle
category: ai
language: en
confidence: high
last_verified: "2026-05-24"
created_date: "2026-05-24"
generation_method: ai_assisted
ai_models:
  - claude-opus
derived_from_human_seed: true
conflict_of_interest: none_declared
is_live_document: false
data_period: static
atomic_facts:
  - id: f1
    statement: >-
      The EU AI Act (Regulation 2024/1689) is the world's first comprehensive AI regulatory framework, using a risk-based approach: unacceptable risk (banned), high-risk (strict compliance), limited
      risk (transparency), minimal risk (no regulation).
    source_title: "European Union. Regulation (EU) 2024/1689: The Artificial Intelligence Act. 2024"
    source_url: https://artificialintelligenceact.eu/
    confidence: high
  - id: f2
    statement: The US Executive Order on AI (October 2023) required developers of powerful AI systems to share safety test results with the government, while NIST developed the AI Risk Management Framework.
    source_title: NIST. AI Risk Management Framework 1.0. 2023-2024
    source_url: https://www.nist.gov/itl/ai-risk-management-framework
    confidence: high
  - id: f3
    statement: The Bletchley Declaration (2023, UK AI Safety Summit) was the first international agreement on frontier AI safety, signed by 28 countries including the US, UK, EU, and China.
    source_title: UK Government. The Bletchley Declaration on Frontier AI Safety. 2023
    source_url: https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration
    confidence: high
completeness: 0.9
primary_sources:
  - title: NIST AI Risk Management Framework 1.0
    type: standard
    year: 2023
    url: https://www.nist.gov/itl/ai-risk-management-framework
    institution: NIST
  - title: The Bletchley Declaration by Countries Attending the AI Safety Summit
    type: official_report
    year: 2023
    url: https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration
    institution: UK Government
known_gaps:
  - Liability frameworks for autonomous AI decisions
  - AI safety research funding mechanisms
disputed_statements:
  - statement: No major disputed statements identified
secondary_sources:
  - title: "The EU AI Act: A Framework for Collaborative Governance of Artificial Intelligence"
    type: journal_article
    year: 2024
    authors:
      - multiple
    institution: Internet Policy Review (ScienceDirect)
    url: https://doi.org/10.1016/j.clsr.2024.106021
  - title: "Global AI Law and Policy Tracker: EU, US, UK, China (IAPP)"
    type: report
    year: 2025
    authors:
      - IAPP Research
    institution: International Association of Privacy Professionals
    url: https://iapp.org/resources/article/global-ai-legislation-tracker
  - title: "AI Regulation Compared: EU AI Act, US Executive Order, UK AI Framework, China AI Law (2026)"
    type: report
    year: 2025
    authors:
      - Legalithm Research
    institution: Legalithm
    url: https://www.legalithm.com/en/blog/ai-regulation-comparison-eu-us-uk-china-global
  - title: "The New Rules of AI: A Global Legal Overview — EU, US, UK, Asia Compliance Risks (Morgan Lewis)"
    type: report
    year: 2025
    authors:
      - Morgan Lewis Research
    institution: Morgan Lewis & Bockius LLP
    url: https://www.morganlewis.com/pubs/2025/12/the-new-rules-of-ai-a-global-legal-overview
updated: "2026-05-24"
---
## TL;DR
AI governance has transitioned from voluntary principles to binding regulation and international cooperation. The Bletchley Declaration, NIST AI RMF, and EU AI Act represent landmark milestones in the emerging global AI governance architecture.

## Core Explanation
Governance layers: (1) International — Bletchley Declaration, Hiroshima AI Process (G7), UN Advisory Body on AI; (2) National — US Executive Orders, China's AI regulations, UK AI Safety Institute; (3) Industry — Frontier Model Forum (Anthropic, Google, Microsoft, OpenAI), voluntary safety commitments; (4) Technical — red teaming, model evaluations, safety cases.

## Detailed Analysis
The AI Safety Institute (UK) and US AISI conduct pre-deployment safety testing. The International Scientific Report on Advanced AI Safety (2024, mandated by Bletchley) synthesizes global research. Key debate: open-source models — should frontier capabilities be openly released?

## Further Reading
- Stanford HAI: AI Policy Tracker
- AI Safety Institutes (UK/US)
- OECD AI Policy Observatory