Chronicle/Policy & standards

Google DeepMind Frontier Safety Framework v3 retains Cyber Uplift Level 1 critical capability level

PolicyFrameworkSignificance assistant-drafted

Google DeepMind's Frontier Safety Framework version 3.0 adds a harmful manipulation CCL and expands misalignment and internal-deployment provisions. Its cyber domain keeps a single CCL, Cyber uplift level 1, for models providing sufficient uplift with high-impact cyber attacks to add expected harm at severe scale, paired with Security Level 2; the framework reasons that automated cyber-defense and social adaptation make higher security levels likely unwarranted.

Why it matters

It shows a lab explicitly reasoning that cyber-capable model weights warrant only Security Level 2, the same as its CBRN and manipulation CCLs and below its ML R&D CCLs, a choice revisited in 2026.

Key facts

As stated in the sources, with where to find them.

  • Cyber CCL 'Cyber uplift level 1': provides sufficient uplift with high impact cyber attacks for additional expected harm at severe scale; recommended Security Level 2.Section 2.2.2, Table 2.2.2.a
  • Rationale: cyber-capable models may interest well-resourced state actors, but automated cyber-defense and social adaptation mean higher security levels are likely not warranted.Table 2.2.2.a, rationale

Findings that cite this record

No tracked finding cites this record yet.

Sources

Related records