Skip to main content
v2026.11,986 entries · CC-BY 4.0
CASRAIAsk CASRAI · AI assistant for researchAsk a research question. Answers come with their sources.Funder rules, compliance, methods and publishing. Answers link the sources they rest on.
NIKOLAI elementN1 · Actors, models and scopeProposednikolai-v0.3

Coverage scope threshold

Definition

A coverage scope threshold is the test that decides whether a safety framework, law or programme applies to a given model or developer. Common tests use the computing power spent on training, the developer's revenue, or a capability benchmark.

Also called: scope of coverage, frontier model definition

NIKOLAI's proposed operational definition

NIKOLAI proposal (unsourced): the if-then test that decides whether a framework, statute or programme applies to a given developer or model (for example, a compute or revenue figure that gates coverage). At least six incompatible scope tests are in active use across this corpus, so NIKOLAI should record which specific test a given claim was produced under rather than treating "in scope" as a single boolean.

The operational definition is CASRAI's own proposal, not a definition any named organisation has agreed to. See what NIKOLAI is and is not, or browse the AI safety glossary.

Examples

Examples from published documents

  • California SB 53 defines a frontier model as "a foundation model that was trained using a quantity of computing power greater than 10^26 integer or floating-point operations" (California SB 53, 22757.11(i))
  • A general-purpose AI model is presumed to have high-impact capabilities "when the cumulative amount of computation used for its training measured in floating point operations is greater than 10^25" (EU AI Act, Article 51(2))
  • Meta's framework applies to "Frontier AI" judged by two criteria, high capabilities in catastrophic risk areas and a compute threshold of at least 10^26 operations (Meta Advanced AI Scaling Framework v2, Appendix I, p.41)

Counter-examples

Not to be confused with

  • A capability threshold decides which safeguards a model needs once it is already in scope; a coverage scope threshold decides whether the framework applies to the model at all.
  • A severity threshold sets how large a harm must be to count as catastrophic; it describes the harm, not which models or developers are covered.

Related terms

Related AI safety terms

Terms defined under this element that are not NIKOLAI elements themselves, each as the cited document uses it. Every term is also listed in the AI safety glossary.

Frontier model
Frontier model, in California SB 53, means a foundation model trained using more than 10^26 integer or floating-point operations, counting the original training run plus later fine-tuning, reinforcement learning and other material modifications.
Source: California SB 53
Frontier AI
Frontier AI, in the Seoul Frontier AI Safety Commitments, means highly capable general-purpose AI models or systems that can perform a wide variety of tasks and match or exceed the capabilities of the most advanced models.
Source: Frontier AI Safety Commitments, AI Seoul Summit 2024
GPAI model with systemic risk
A general-purpose AI model with systemic risk is the EU AI Act class with the heaviest model duties. A model qualifies if it has high-impact capabilities, presumed above 10^25 training FLOP, or if the Commission designates it.
Source: EU AI Act Article 51
High-impact capabilities
High-impact capabilities, under EU AI Act Article 3(64), are capabilities that match or exceed those recorded in the most advanced general-purpose AI models. Having them is the main route to systemic-risk classification under Article 51.
Source: EU AI Act Article 3 (Definitions)
Compute threshold
A compute threshold is a training-compute figure, counted in floating-point or integer operations, above which a law or framework applies. The EU AI Act presumes high-impact capabilities above 10^25 FLOP; California SB 53 uses 10^26 operations.
Source: EU AI Act Article 51
Covered model
A covered model is a model that falls inside a framework's or law's scope. OpenAI's Frontier Governance Framework covers frontier models under California's TFAIA and general-purpose models with systemic risk under the EU AI Act.
Source: OpenAI Frontier Governance Framework

Source of record

Where this definition comes from

  • Meta Advanced AI Scaling Framework v2, Appendix I, p.41

    “"Frontier AI" has two criteria: "High Capabilities in Catastrophic Risk Areas" and "Compute Threshold: We trained the model using at least 10^26 integer or floating point operations (to include material modifications ...)"”

    https://ai.meta.com/static-resource/Meta_Advanced-AI-Scaling-Framework-v2
  • California SB 53, 22757.11(i)

    “"Frontier model": "a foundation model that was trained using a quantity of computing power greater than 10^26 integer or floating-point operations", including "any subsequent fine-tuning, reinforcement learning, or other material modifications"”

    https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53
  • EU AI Act Article 51, Article 51(1)-(3)

    “Article 51(1): "(a) it has high impact capabilities evaluated on the basis of appropriate technical tools and methodologies" or "(b)" a Commission designation decision; Article 51(2): presumed high-impact "when the cumulative amount of computation used for its training measured in floating point operations is greater than 10^25", rebuttable.”

    https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-51

Crosswalk

How named organisations use this concept

Every row below is a shadow mapping. A shadow row is CASRAI's own reading of a published document. No lab, evaluator or regulator named on a shadow row has declared, endorsed, or been consulted on it. That changes only when an organisation files its own Mapping Declaration.
OrganisationTheir term, as publishedMatch & verificationSource
AnthropicShadow mapping
Advanced AI Framework
“Covered Developer test (>10^25 training FLOP or >$500M annual AI-derived revenue or >$1B AI R&D spend); and "Over time, the FLOP needed to train dangerous models may fall, and it may make sense to introduce a threshold based on capabilities rather than simply on training costs." (AAF p.3)”closeCL
confidence: medium
Anthropic Advanced AI Framework, legislative proposal
OpenAIShadow mapping
Frontier Governance Framework / Preparedness Framework v2
“"frontier models as defined under the TFAIA and 'general-purpose models with systemic risk' under the EU AI Act" (FGF §1). PF: "Covered deployment": "any new or updated deployment that has a plausible chance of reaching a capability threshold whose corresponding risks are not addressed by an existing Safeguards Report" (§3.2).”closeCL
confidence: medium
OpenAI Frontier Governance Framework
Google DeepMindShadow mapping
Frontier Safety Framework v3.1
“"Frontier AI Models or Models: are trained on a large data set, display significant generality, are capable of performing a wide range of distinctive tasks and have high-impact capabilities. Frontier AI models' agentic and reasoning-based general capabilities near or exceed those of other Google models." (glossary, p.18). This is a relative test, not an absolute number.”closeCL
confidence: medium
Google DeepMind Frontier Safety Framework v3.1
xAIShadow mapping
Frontier AI Framework, 30 Jun 2026
“"xAI's frontier AI models, such as Grok" (s.1); no compute or capability criterion given at all.”
This source's PDF metadata /Title reads "Privileged/Confidential DRAFT working FRAMEWORK DOC"; no xAI statement disambiguating draft vs. final status was found, so its status remains unresolved.
noneDU
confidence: medium
xAI Frontier AI Framework, 30 Jun 2026
MetaShadow mapping
Advanced AI Scaling Framework v2
“"Frontier AI" has two criteria: "High Capabilities in Catastrophic Risk Areas" and "Compute Threshold: We trained the model using at least 10^26 integer or floating point operations (to include material modifications ...)" (Appendix I, p.41).”exactEQ
confidence: high
Meta Advanced AI Scaling Framework v2
EUShadow mapping
AI Act Article 51 / Safety and Security chapter
“Article 51(1) sets two alternative tests: "(a) it has high impact capabilities evaluated on the basis of appropriate technical tools and methodologies, including indicators and benchmarks"; or "(b) based on a decision of the Commission ... having regard to the criteria set out in Annex XIII." Article 51(2): "A general-purpose AI model shall be presumed to have high impact capabilities pursuant to paragraph 1, point (a), when the cumulative amount of computation used for its training measured in floating point operations is greater than 10^25", rebuttable, with Article 51(3) empowering the Commission to amend the threshold by delegated act.”
No "10^23" figure appears anywhere in Article 51. A "10^23 GPAI indicator" has not been located in this source; CASRAI has not verified it against Annex XIII or the Commission's July 2025 Guidelines, so it is not used here as an EU coverage-scope test. Article 3's definitions were read directly from the AI Act Service Desk: Article 3(64) "'high-impact capabilities' means capabilities that match or exceed the capabilities recorded in the most advanced general-purpose AI models" and Article 3(65) defines "systemic risk".
exactEQ
confidence: high
EU AI Act Article 51
California SB 53Shadow mapping
SB 53 statute text
“"Frontier model": "a foundation model that was trained using a quantity of computing power greater than 10^26 integer or floating-point operations", including "any subsequent fine-tuning, reinforcement learning, or other material modifications" (22757.11(i)).”exactEQ
confidence: high
California SB 53
US Government (Executive Order 14409)Shadow mapping
Executive Order 14409
“"a classified benchmarking process to assess the advanced cyber capabilities of AI models and determine the threshold at which an AI model should be designated a 'covered frontier model'" (Sec. 3(a)). Cyber-only and classified.”narrowNR
confidence: medium
Executive Order 14409
Demis Hassabis (personal essay)Shadow mapping
"A framework for frontier AI" essay
“"A model would qualify as 'Frontier-class' if it meets certain thresholds on a set of benchmarks determined by the Standards Body and regularly updated" (§Framework ¶2).”
Personal essay, not an official DeepMind policy document.
closeCL
confidence: medium
Demis Hassabis, "A framework for frontier AI and the dawning of a new age"
MicrosoftShadow mapping
Frontier Governance Framework (Microsoft)
“Scopes its leading-indicator assessment to models in scope under the EU AI Act, TFAIA and RAISE, and to substantial fine-tunes over one-third of base compute (para.).”closeCL
confidence: medium
Microsoft Frontier Governance Framework
US Congress (H.R. 9925, FRONTIER Act, not enacted)Shadow mapping
FRONTIER Act bill text
“Sets a tiered test: "frontier developer" at >10^26 operations; "large frontier developer" additionally requires >$50M revenue and >=$1B in AI-related development expenditure over a rolling 36-month window; "very large frontier developer" additionally requires >$5B revenue and >=$10B in AI spending -- each tier inheriting the duties of the tiers below it (Sec. 2, Sec. 6).”
Bill not enacted.
closeCL
confidence: medium
FRONTIER Act, 119th Congress (H.R. 9925, not enacted)
What do these codes mean?
exact
The source term is equivalent to this element
close
The source term is close but not equivalent to this element
broad
The source term is broader than this element
narrow
The source term is narrower than this element
none
No mapping claim — used for false-friend and declared-but-undefined rows
EQ
Equivalent
CL
Close
BR
Source is broader than the element
NR
Source is narrower than the element
FF
False friend — same or similar label, different meaning
DU
Declared but undefined by the source
UV
Unverified

Divergence

Where sources materially disagree

"Frontier model" is itself a false-friend label across this corpus: 10^26 FLOP (SB 53, Meta); 10^25 FLOP plus revenue (Anthropic AAF); relative to Google's own models (GDM); a classified cyber benchmark (EO 14409); a benchmark set refreshed by a standards body (Hassabis). Critically, the EU AI Act itself never uses the phrase "frontier model" at all -- its gate is the distinct "general-purpose AI model with systemic risk" classification under Article 51, with its own two-part test. NIKOLAI should treat "frontier model" and "GPAI model with systemic risk" as related but non-identical scope concepts, not as translations of one another. "Systemic risk" is also a term of art specific to the EU chapter and is not the same scope as SB 53's "catastrophic risk" (50 deaths/$1B, single-incident) or the labs' "severe harm"; xAI's FAIF borrows the EU's four-domain taxonomy but not its legal test.

Gap

Six incompatible scope tests are in use (10^25 FLOP plus revenue; 10^26 FLOP; relative to the developer's own models; a classified cyber benchmark; a benchmark set that a standards body would refresh; and H.R. 9925's three-tier compute-plus-revenue-plus-spending ladder). A risk report cannot be compared across regimes without recording which test it was produced under.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →