What Is Systemic Risk (GPAI)?
Systemic Risk (GPAI) is under the EU AI Act, the category of risk posed by the most capable general-purpose AI models whose impact could propagate across the economy, society, or fundamental rights at scale.
Systemic Risk (GPAI) — under the EU AI Act, the category of risk posed by the most capable general-purpose AI models whose impact could propagate across the economy, society, or fundamental rights at scale.
The EU AI Act presumes a GPAI model poses systemic risk if it was trained using more than 10²⁵ floating-point operations (FLOPs) — a compute threshold intended to capture frontier models. Providers of systemic-risk GPAI face additional obligations: model evaluations and adversarial testing, systemic-risk assessment and mitigation, serious-incident reporting, and cybersecurity protection. The category was among the most heavily negotiated parts of the Act and can be adjusted by the Commission as the technology evolves.
Source: EU AI Act, Articles 51–55
Plain-language explanation
The EU AI Act presumes a GPAI model poses systemic risk if it was trained using more than 10²⁵ floating-point operations (FLOPs) — a compute threshold intended to capture frontier models. Providers of systemic-risk GPAI face additional obligations: model evaluations and adversarial testing, systemic-risk assessment and mitigation, serious-incident reporting, and cybersecurity protection. The category was among the most heavily negotiated parts of the Act and can be adjusted by the Commission as the technology evolves.
See where you stand on AI governance
Take the free 7-question maturity assessment and get a personalised action plan.
Free assessment — 3 minutes →