Every reviewed AI governance resource in one place — reports, standards, guidance and regulation.
Press release from the U.S. Bureau of Industry and Security announcing a regulatory framework of export controls on advanced computing chips and closed AI model weights, with new licence exceptions, Data Center Validated End User updates and security conditions.
Legal advisory from the California Attorney General's Office explaining how existing California consumer protection, civil rights, competition, privacy and election laws apply to AI, with a companion advisory on AI in healthcare and new AI statutes effective 1 January 2025.
A research report identifying Chinese "AISI counterparts" — government-linked institutions doing work similar to the US and UK AI Safety Institutes — and assessing their suitability for engagement on evaluations, standards and cooperation.
An IHEID report on the geopolitical implications of the race to develop quantum computers, covering research collaboration, intellectual property, protectionist industrial policies, pre- and post-quantum risks, and governance recommendations.
UNECE WP.6 publication setting out an overarching common regulatory arrangement for products and services with embedded AI, covering regulatory objectives, risk assessment, standards, conformity assessment and market surveillance, plus a declaration for government agencies.
Deloitte Legal white paper on contractual issues and mitigations for organisations procuring generative AI systems and for GenAI use by suppliers and service providers, including EU AI Act implications.
A governance framework, sponsored by Connor Group, that divides generative AI governance into five domains and pairs named risk categories with control considerations, plus a four-step adoption approach and a five-stage risk assessment.
U.S. Copyright Office report, Part 3 of its Copyright and Artificial Intelligence study, analysing how copyrighted works are used in generative AI training and how fair use and licensing apply, released in a pre-publication version.
A 28-page PDF whose extracted text contains only a table of contents and running page numbers. The visible headings address cross-regulatory cooperation and EU data protection law as a 'cornerstone'; publisher, date and authorship are not recoverable.
An IAPP reference matrix mapping the articles of the EU AI Act to the operators they primarily concern — providers, deployers, importers, distributors, product manufacturers and authorised representatives — across high-risk AI systems, AI systems and general-purpose AI models.
DIGITALEUROPE manifesto for Europe in 2030: an assessment of EU digital progress against its 2025 KPIs and Digital Decade data, its own KPIs for 2030, and 20 policy solutions on the single market, data, cyber, standards, skills and AI.
A law firm guide to the EU Artificial Intelligence Act, published by Bird & Bird and dated 6 November 2024, working through scope, prohibited practices, high-risk systems, general-purpose AI models, transparency, sandboxes, enforcement and upcoming secondary legislation.
A July 2024 report authored by Deloitte and International Organisation of Employers staff assessing G20 readiness for AI: labour-market effects, investment, infrastructure and skills indicators, company case studies and national AI strategies.
A workshop-driven report that translates broad control language from standards and policies into fragmented, testable control activities and test procedures, with templates, facilitation scripts and sample wordings and test designs for twelve control types.
FS-ISAC white paper explaining the purpose, methodology and use of its Generative AI Vendor Risk Assessment workbook, which helps financial institutions tier due diligence and assess GenAI vendors through risk analysis, questionnaires and a final report.
World Economic Forum insight report, in collaboration with Capgemini, introducing the Agent Capability and Authorization Profile (ACAP) as a deployment-level authorization instrument, alongside enterprise agent guidelines and a three-phase adoption life cycle.
Anthropic's Responsible Scaling Policy version 3.0 is a voluntary framework describing how the company identifies, evaluates and mitigates catastrophic risks from advanced AI, including industry-wide safety recommendations, Frontier Safety Roadmaps, Risk Reports and governance commitments.
Google DeepMind's Frontier Safety Framework v1.0 sets out protocols for detecting Critical Capability Levels in Autonomy, Biosecurity, Cybersecurity and ML R&D, and pairs each risk domain with graded security and deployment mitigations.
World Economic Forum white paper, produced with CPP Investments Insights Institute, offering investors a three-step playbook for responsible AI: internal commitments, portfolio due diligence, and engagement with companies, asset managers and the wider ecosystem.
Anthropic's Responsible Scaling Policy version 1.0 sets out AI Safety Levels that tie increasing model capability to stricter security, evaluation and deployment measures, with initial commitments defined for ASL-2 and ASL-3.
A technical security guide for designing, developing, deploying, testing, and operating LLM-powered agentic applications. It provides architecture-specific controls, lifecycle practices, operational safeguards, and assurance approaches.
OWASP's 2025 guidance lists ten security vulnerabilities in LLM applications. Each entry defines the risk and provides examples, prevention and mitigation strategies, attack scenarios, and reference links.
This report surveys agentic AI security and governance risks, documented incidents, protocols, identity, supply-chain provenance and regulation. It provides an agent taxonomy and a maturity model linking adoption tiers to governance capability.
A 2026 OWASP Top 10 guide to security risks in LLM-integrated applications. It defines ten risk categories, mitigation strategies, attack scenarios, and mappings to related security frameworks and taxonomies.
Curated Library of AI Governance Resources
