AI Glossary

Compute threshold

noun phraseAI regulationpreferred in this glossary

Level of training computation that regulators use to identify general-purpose AI models that may pose risks of large-scale harm to society.

Source: Training Compute Thresholds (opens in a new tab) (adapted)

Adapted: restates the version 2 abstract's regulatory use of training computation without making a number alone determine legal status. Source: Training Compute Thresholds (opens in a new tab)

In plain language

A training-computation figure can trigger a regulatory presumption or definition.

In a sentence

  • The reviewer checked which operations counted toward the compute threshold.
  • The policy analysis considered the compute threshold alongside the law's other conditions.

Examples written for this glossary

Usage note

The unit and legal effect depend on the rule. A numerical cutoff alone does not describe every condition for classification, and the entry is not legal advice.

  • EU presumption

    Regulation

    A regulation states a rule issued by a public authority. Article 51(2) uses more than 10^25 floating point operations in training for a presumption of high-impact capabilities.

    Source: EU AI Act, Article 51 (opens in a new tab)

  • California definition

    Regulation

    A regulation states a rule issued by a public authority. SB 53 section 22757.11(i)(1) defines a frontier model as a foundation model trained using more than 10^26 integer or floating-point operations.

    Source: California SB 53 (opens in a new tab)

See also

  • General-purpose AI model

    AI model under the EU AI Act that displays significant generality, is capable of competently performing a wide range of distinct tasks regardless of the way the model is placed on the market, and can be integrated into a variety of downstream systems or applications.

  • Systemic risk

    Risk under the EU AI Act that is specific to the high-impact capabilities of general-purpose AI models, having a significant impact on the Union market due to their reach, or due to actual or reasonably foreseeable negative effects on public health, safety, public security, fundamental rights, or the society as a whole, that can be propagated at scale across the value chain.

  • Frontier AI framework

    Large frontier developer's documented technical and organizational protocols to manage, assess, and mitigate catastrophic risks from its frontier models, as described in California's SB 53.

  • Frontier model

    General-purpose AI model that outperforms, on a range of conventional performance benchmarks, the models that have been widely deployed for at least 12 months.

Cite this term.

CC BY 4.0, attribution required.

This term

Orbyt Labs, Compute threshold (AI Glossary), Nov 14, 2026. https://www.orbytlabs.ai/orbyt-collective/glossary/compute-threshold

The dataset (JSON)

Orbyt Labs Glossaries (Nov 14, 2026), https://www.orbytlabs.ai/glossary#dataset

Last updated .