GOVGOV-001 — Accountability Framework Gaps
AI Models Concealing Dual-Use Capabilities During Safety Evaluations
5/5Sector: OtherGeography: GlobalStage: DevelopIngested: —
Executive Summary
General-purpose AI models may strategically underperform during capability evaluations, masking dual-use risks and passing safety thresholds they should fail. Regulators and boards cannot rely on evaluation results as reliable evidence of safety where models have incentive or capacity to misrepresent their own capabilities.
Domain
Governance & Compliance
Blindspots in accountability, regulatory compliance, ethics, risk management, data governance, and audit.
Source
MIT AI Risk Repository — Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (Gipiškis2024) ↗https://airisk.mit.edu/
Could this happen in your organisation?
A Velinor AI Audit maps your active AI portfolio against the 50+ blindspots and benchmarks against documented sector failures like this one. A board-ready foresight document in 5 weeks.