AI Governance Engineer
GitHub

Critical Capability Level (CCL)

By Updated

In Google DeepMind's Frontier Safety Framework, a capability level at which a model may pose a heightened risk of severe harm without mitigations; reaching one calls for security and deployment mitigations and a decision that the residual risk is acceptable. Version 3.1 adds lower Tracked Capability Levels 1.

Developed in
the frontier safety frameworks crosswalk
Chapters
ch. 10, Reading List
Source
1 numbered reference, listed below

Where it is used

The term is not used under this name in running prose; the sections listed under "Developed in" treat it.

Critical Capability Level (CCL): related terms and patternsThe term at the centre; related terms on solid edges, the terms it is contrasted with on dashed edges, and the patterns that use it most in the colour of their layer.RelatedCritical Capability Level (CCL)Related terms (3)AI Safety Level (ASL) · RelatedAI Safety Level (ASL)Frontier Safety Framework · RelatedFrontier Safety FrameworkResidual risk · RelatedResidual risk
Critical Capability Level (CCL): related terms and patternsThe term at the centre; related terms on solid edges, the terms it is contrasted with on dashed edges, and the patterns that use it most in the colour of their layer.RelatedRelated terms (3)AI Safety Level (ASL) · RelatedAI Safety Level (ASL)Frontier Safety Framework · RelatedFrontier Safety FrameworkResidual risk · RelatedResidual riskCritical CapabilityLevel (CCL)
The term's neighbourhood Terms related to this one on solid edges, terms it is often confused with on dashed edges, and the patterns that use it most in the colour of their layer. Follow a dashed edge to check the distinction before you use either term.
Neighbourhood as a table
Neighbourhood of Critical Capability Level (CCL)
FamilyItemRelation
Related termsAI Safety Level (ASL)Related
Related termsFrontier Safety FrameworkRelated
Related termsResidual riskRelated
  • AI Safety Level (ASL)

    Anthropic's name for tiers of safeguards in its Responsible Scaling Policy.

  • Frontier Safety Framework

    Google DeepMind's published policy for severe risks from its most capable models: Critical Capability Levels for misuse, machine learning R&D and…

  • Residual risk

    What is left of a risk once treatment is applied.

Sources

  1. [1] Frontier Safety Framework, Version 3.1 (Critical and Tracked Capability Levels, security and deployment mitigations, annual review). Google DeepMind. 2026-04-17. https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf (verified: primary)

Definitions of legal terms paraphrase the cited text, which governs. Dated statements are as of .

Cite this term

García Aibar, J. (2026). Critical Capability Level (CCL). In AI Governance Engineering: The Thesis & Body of Knowledge (v0.5.0), Glossary. https://doi.org/10.5281/zenodo.22956197. https://aigovernanceengineer.com/glossary/critical-capability-level-ccl. CC BY 4.0

BibTeX

@misc{aige2026criticalcapabilitylevelccl,
  author  = {Jorge García Aibar},
  title   = {{Critical Capability Level (CCL)}},
  note    = {Glossary, AI Governance Engineering: The Thesis \& Body of Knowledge, version 0.5.0},
  year    = {2026},
  doi     = {10.5281/zenodo.22956197},
  url     = {https://aigovernanceengineer.com/glossary/critical-capability-level-ccl}
}