Governance
GPAI provider with systemic risk
Provider whose model can cause systemic societal or rights harms.
Definition
A provider whose GPAI model is capable of causing systemic societal, safety, security or fundamental-rights harms; such providers and their models are subject to additional Safety & Security measures in the Code, including adversarial testing, red-team exercises, continuous monitoring and enhanced incident reporting.
Related Terms
Systemic-risk GPAI model
GPAI model designated as posing systemic risks to the Union....
Unsafe AI System
AI posing unreasonable risks to health, safety, or rights....
High-risk system
AI use posing substantial effects on rights, health or essential services....
High‑risk AI
AI uses with potential for severe or large‑scale harm....
High‑risk system
Uses posing substantial risk to rights, safety, or infrastructure....