Corporate AI Risk Assessment Framework (Draft)

Partnership on AI published a draft Corporate AI Risk Assessment Framework designed to move companies "from a nebulous and theoretical approach to AI risk management to one that is precise, practical, and actionable." The framework is intended for use at the corporate level — not the AI system level — covering a company's entire upstream and downstream value chain as well as its own operations. It consolidates risks across PAI's areas of expertise including labor impacts, transparency, and synthetic media, and is structured to inform formal financial and sustainability reporting, enterprise risk management, materiality assessments, and human rights due diligence. PAI is seeking feedback on practical use, topic coverage, and how the framework can sharpen investor due diligence. The release addresses documented industry feedback that existing resources from ISO, NIST, and OECD are valuable but lack the precision needed for corporate implementation.

This is the first major framework explicitly designed for board-level and enterprise risk management rather than technical AI system assessment. As AI deployments scale, CROs and compliance teams need structured tools that integrate AI risk into existing corporate governance processes. The framework's value-chain scope addresses a documented gap: most AI governance focuses on models or products, but enterprise risk materializes across adoption, procurement, and partnerships.

Chief Risk Officers and enterprise AI governance leads should review the draft framework and provide feedback to PAI by mid-June — early engagement shapes the final tool's applicability to your sector. Organizations currently using ISO 42001 or NIST AI RMF should assess whether this corporate-level complement addresses gaps in their materiality and disclosure practices.