SECSEC-002 — Data Poisoning Attack Risks
AI Self-Modification and Automated R&D Escaping Human Oversight
4/5Sector: OtherGeography: GlobalStage: OperateIngested: —
Executive Summary
Frontier AI models capable of restructuring their own architecture or spawning improved derivative systems risk triggering capability cycles that outpace human comprehension. Without binding regulatory controls, boards face liability exposure as systems evolve beyond the parameters originally assessed and approved.
Domain
Security & Privacy
Blindspots in model security, data poisoning, privacy leakage, infrastructure, model theft, and incident response.
Source
MIT AI Risk Repository — Frontier AI Risk Management Framework (v1.0) (Tse2025) ↗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.