Compute Threshold
Threshold of computational resources (FLOPS) used to classify AI models for additional regulatory requirements.
Definition
Compute Threshold is a quantitative metric used in AI regulation to classify AI models based on the computational resources used in their training, triggering enhanced regulatory requirements.
EU AI Act Threshold (Article 51):
- 10^25 FLOPS: A general-purpose AI model is presumed to have high impact capabilities (and thus systemic risk) when the cumulative amount of compute used for its training is greater than 10^25 floating point operations
- Measured in: Floating point operations (FLOPS)
- Adjustable: The Commission can update this threshold through delegated acts based on technological developments
Rationale for Compute Thresholds:
- Proxy for Capability: Computational scale correlates with model capabilities
- Measurable: More objective than capability-based assessments
- Predictable: Allows providers to know classification in advance
- Scalable: Can be adjusted as technology evolves
Current Context (2024):
- GPT-4 estimated at ~10^25 FLOPS (near threshold)
- Newer frontier models likely exceed this threshold
- Threshold roughly corresponds to current "frontier" models
Obligations When Exceeding Threshold:
- Enhanced model evaluation and testing
- Systemic risk assessment and mitigation
- Serious incident tracking and reporting
- Adequate cybersecurity protections
- Cooperation with authorities
Limitations:
- Does not account for training efficiency improvements
- Same compute can produce different capability levels
- Does not address fine-tuned or open-weight models
- May become outdated as technology advances
Sources
- •EU AI Act Article 51
- •Executive Order 14110
Related Terms
General-Purpose AI Model with Systemic Risk
A general-purpose AI model legally designated as carrying risks that can affect the EU market or society at scale, subject to heightened obligations under the EU AI Act....
Systemic Risk
Risk that a general-purpose AI model could cause widespread, significant harm to public health, safety, security, or society....
Frontier AI
Highly capable, cutting‑edge AI models (a subset of general‑purpose/foundation models) whose scale or capabilities may create systemic or catastrophic risks....