A Practical Guide for Technical Leaders: AI Governance
Effective AI governance, however, is not merely about placing limits on models; it is about ensuring the integrity, accountability, and security of the data ecosystems that power them.
AI Governance Library: all resources
Every reviewed AI governance resource in one place — reports, standards, guidance and regulation.
State of AI Safety in China (2026)
Over the three years this report series has tracked, it has become increasingly clear that China’s approach to AI safety and governance extends far beyond content control. With the rise of autonomous agents, the concern has expanded to what AI does and what it does to people and society.
An AI Policy Is Not AI Governance: Seven Imperatives to Govern AI
If you don't manage AI risk properly, one of two things happen. Either people innovate wildly and you're exposed – or they stop innovating because they don't feel protected, or it's too hard. Either way, you lose.
Prysmian Group AI Policy: Corporate AI Governance Framework
This Policy outlines Prysmian’s commitment to responsible AI use to ensure ethical considerations are upheld, AI risks are managed, and compliance to applicable regulations is achieved. The document provides a framework to guide all AI-related activities, establishing acceptable behaviors.
Standards as a Basis for the Global Governance of AI in Research
The research sector is poised to be a high-risk but low-governance space, the worst combination of circumstances in which AI harms could proliferate globally.
AIVSS Scoring System for OWASP Agentic AI Core Security Risks v0.8
The framework posits that agentic capabilities do not just add new risks; they fundamentally expand the "blast radius" of existing technical vulnerabilities.
The Rise in Unstructured Data and AI Security Risks
Only 35% of organizations report full visibility into where unstructured data resides, while classification and labeling practices remain inconsistent. Security, governance, privacy, and compliance are top concerns, yet many organizations struggle to execute foundational controls.
The Implications of Artificial Intelligence in Cybersecurity: Shifting the Offense-Defense Balance
The report concludes that in the near term, AI offers a significant advantage to cyber defenders, particularly those who can capitalize on their "home field" advantage and first-mover status. However, sophisticated threat actors are also leveraging AI to enhance their capabilities.
AIMS Simplified: What ISO 42001 Actually Requires
38 controls. 8 domains. One governance architecture. The organisations that treat these as a documentation exercise will not pass audit. The organisations that treat them as an operating model will.
AI Information Threats and Crisis Response: Practitioners’ Handbook
AI information threats are exacerbating risks to democracy and real-world violence during crisis events – such as terrorist attacks, and national or international security incidents.
A Pragmatic Approach to Regulating AI Agents: EU AI Act Classification and Contract Law Guardrails
While agents generally utilise general-purpose AI models, we posit that their structural complexity and cross-system permeability require them to be regulated as "AI systems" with distinct obligations under the AI Act. Consequently, our proposals highlight the need for robust accountability.
ISO/IEC 42001:2023 AIMS Clause-Wise Implementation Checklist
ISO/IEC 42001:2023 is the world’s first international standard for establishing, implementing, maintaining, and continually improving an Artificial Intelligence Management System (AIMS).
Enterprise AI Security Starts With AI Agents
The risk associated with AI agents is no longer theoretical. Survey findings indicate that autonomous systems are already exceeding intended permissions and operating outside intended scope in day-to-day efforts. In many environments, these behaviors occur at least occasionally.
AI RMF 2026 Implementation Manual: Aligning IEEE AI Governance, Risk, and Safety Standards
This chapter applies to AI systems that create, support, recommend, rank, classify, generate, optimize, or autonomously execute decisions, actions, or content where a human user, impacted individual, operator, customer, or regulator may reasonably require an understandable account of system behavior
Careful Adoption of Agentic AI Services
Until security practices, evaluation methods and standards mature, organisations should assume that agentic AI systems may behave unexpectedly and plan deployments accordingly, prioritising resilience, reversibility and risk containment over efficiency gains.
Sound Practices for Responsible Adoption of Artificial Intelligence (AI)
The 12 sound practices cover organisation-wide governance as well as management of different stages of AI development and deployment (or AI lifecycle). When applying the sound practices, financial institutions should consider proportionality.
Opportunities for AI in Cyber Defence
AI can significantly enhance cyber security, but it is not a replacement for strong cyber security fundamentals. Poorly designed or poorly governed AI systems can introduce new attack paths.
Governing Agents: A Practitioner's Cross-Framework Reference
The central claim of this reference is that AI agents are not a variant of AI systems for which existing compliance mappings can be mechanically reused. They are a distinct governance category.
Strategy and Governance Failures: Why AI Programs Fail Before They Start
The single most consequential failure in enterprise AI adoption is not technical — it is the collapse of governance at the point of adoption. Organizations are investing at historic levels in AI technology while simultaneously failing to build the organizational infrastructure that makes those inves
A Formal Model of How AI Erodes Human Agency
In this report, we develop a formal framework for measuring how artificial intelligence (AI) affects collective human agency in decisionmaking processes. We define collective agency as humanity’s capacity to make consequential choices about shared resources, institutions, and the social environment.
AI Vendor Assessment Guide: A Live-Dialogue Evaluation Framework
Applying a standard procurement lens to an AI vendor tends to produce a false sense of assurance. This guide is not a replacement for your existing procurement process. It is a supplement designed for the specific risks that AI vendors introduce.
G7 Software Bill of Materials for AI: Minimum Elements
Drawing from the existing Software Bill of Materials (SBOM) concept, an SBOM for AI consists of a structured record, or inventory of details and supply chain relationships for the various components used in building an AI system.
Responsible AI in Practice: 2025 Global Insights from the AI Company Data Initiative
The data confirms both the immense momentum of AI adoption and the stark reality of the current governance gap. Much like the broader ecosystem, companies across industries are embedding new AI capabilities faster than they are formalising accountability, internal controls, and oversight.
Artificial Intelligence and Machine Learning: Supply Chain Risks and Mitigations
Adopting AI and ML systems introduces unique supply chain risks, which can threaten the cyber security of an organisation if not securely managed. Using pre-trained models and third-party datasets is beneficial, but it can also bring existing compromise and supply chain risks.
