Technical
Model-intrinsic risk
Risk stemming from a model's inherent capabilities or behaviours.
Definition
Risks originating in the model itself—such as emergent capabilities, propensity to generate harmful outputs, or systemic societal hazards—that can pose broader threats beyond any single deployment and may require cross‑regulatory monitoring or targeted mitigation.
Related Terms
Model Risk
Risk of adverse consequences from incorrect or misused model outputs....
Inherently Dangerous AI System
An AI system posing significant risk of severe harm if not properly managed....
AI-specific cybersecurity threats
Security threats unique to AI such as prompt injection and model inversion....
Foundation models / Frontier AI
Highly capable, general-purpose AI models with broad applicability....
Risk profile
Categorization of foreseeable harms by likelihood and severity....