Every reviewed AI governance resource in one place — reports, standards, guidance and regulation.
The guidance defines five categories of agentic AI risk — privilege escalation, design and configuration failures, behavioral misalignment, structural brittleness, and accountability gaps — and requires each agent to carry a verified, cryptographically anchored identity with short-lived credentials.
Trust is no longer a property of a system; it is a managed outcome. Trust emerges from continuous oversight, governance structures, and organisational capabilities rather than from the intrinsic qualities of a model or vendor.
Traditional IAM comprises the identity providers, directories, federation protocols, and access-control systems built for human users and relatively static machine identities. These systems assume long-lived principals, coarse roles, and “authenticate once, trust for the session”
One architectural property makes GenAI data security fundamentally different from every prior computing model: the context window aggregates data from multiple trust domains (system prompt, user input, RAG results, tool outputs, conversation history) into a single flat namespace with no internal acc
This document provides a structured, practical comparison between AI Risk and Information Security Management System (ISMS) Risk, as defined by ISO/IEC 27001:2022. It is intended to support CISOs, Chief Risk Officers, compliance leads, and Board-level decision-makers
Resilience cannot be reduced to personal 'grit' or mindfulness. It must be treated as a civic design imperative and built into the systems and cultures that shape public life. ... We need systems that will protect human agency, not automate it.
Given the pace of AI R&D automation and the limited external visibility into how companies use their most capable models internally, regular and detailed risk reporting may be one of the few mechanisms available to ensure that the risks from internal AI use are identified and managed
This document provides an AI risk-management standards profile (hereafter “Profile”), a targeted set of risk-management practices or controls specifically for identifying, analyzing, and mitigating risks of GPAI models.
A policy report arguing that AI governance must move beyond fragmented national approaches toward a UN-centered global framework built around four pillars: safety, inclusiveness, clarified responsibilities, and multilateral coordination.
Many vendors overclaim… focusing on superficial jailbreak demonstrations… while ignoring systemic vulnerabilities, workflow bypasses, misuse risks, and failures in agentic or tool-enabled systems.
AI is increasingly being used to augment and automate cyber operations… raising urgent questions about when AI systems introduce unacceptable cyber risk, and how risk thresholds should be identified before harms materialize at scale.
The OECD.AI Index… is a composite measurement framework designed to assess countries’ implementation of the OECD Recommendation on Artificial Intelligence.
An ability of individuals, organisations, and systems to perform AI regulatory functions effectively, efficiently, and sustainably across the lifecycle...
Governance is not merely a set of external constraints… but an intrinsic feature of any AI system, adaptable to evolving risks and contexts.
Frontier AI both amplifies existing risks and introduces qualitatively novel challenges… emerging safety practices are often misaligned with established risk management frameworks.
AI actors should carry out risk-based due diligence to identify, prevent, mitigate and account for adverse impacts of AI systems across their lifecycle.
Model Risk Management (MRM) must evolve from periodic validation to continuous governance as Generative and Agentic AI systems introduce opacity, stochasticity, and autonomy that break traditional supervisory assumptions.
The AICM provides a foundational security and governance framework… establishing a clear and actionable set of controls… tailored to the unique risks and responsibilities inherent in AI development, deployment, and consumption.
This Report assesses what general-purpose AI systems can do, what risks they pose, and how those risks can be managed… focusing on emerging risks at the frontier of AI capabilities.
Insurance can help ensure that AI-related harms are mitigated, and that AI’s risks and benefits are fairly shared… by offering risk-based premiums tied to safety standards.
Agents don't just generate content; they act, adapt and influence digital and physical environments across extended time horizons, with compounding and potentially irreversible consequences.
AI is rapidly becoming deeply embedded in contemporary societies… reshaping industries, economies, and societies… governments are establishing policies, legal frameworks and ethical guidelines to ensure responsible use.
This framework aims to help organizations harness the power of generative artificial intelligence (GenAI) while appropriately managing its risks.
A Fundamental Rights Impact Assessment (FRIA) supports identifying, assessing and mitigating risks AI systems may pose to individuals’ rights, ensuring compliance with the EU AI Act and broader human rights standards.
Curated Library of AI Governance Resources
