Google DeepMind

Gemini 3.1 Pro

• Framework: Frontier Safety Framework
• 2 statements about capability levels
• 5 independent or government tests
• 1 case names this version
• 7 tests by the provider
• 11 source documents

Incidents naming this model

1 incident names this model version. Each shows its risk grouping and what was done, as a source states it.

What the provider states

• Chemical, biological, radiological and nuclear · Uplift level 1 · Stated as not crossed

  we found that the model remains below alert thresholds for the cbrn, harmful manipulation, machine learning r&d, and misalignment ccls.

  Google DeepMind · 19 February 2026 · Read the source
• Cyber · Uplift level 1 · Stated as not crossed

  as previous models passed the alert threshold for cyber, we performed more additional testing in this domain on gemini 3.1 pro with and without deep think mode, and found that the model remains below the cyber ccl.

  Google DeepMind · 19 February 2026 · Read the source

Independent and government tests

• per-criterion scoring results · RAND, independent evaluator · Chemical, biological, radiological and nuclear

  the heatmaps of figures d.1 and d.2 show the number of attempts, out of 10, in which gemini 3.1 pro and deepseek v4 pro achieved each of our validity, structure, and function- related scoring criteria.

  RAND (CAST / Project Canary) · 11 August 2026 · Read the source
• SecureBio dual-use biology benchmarks · SecureBio, independent evaluator · Chemical, biological, radiological and nuclear

  gemini 3.1 pro has improved on highly informative samples we find relevant to expert troubleshooting and failure identification.

  RAND (CAST / Project Canary) · 6 August 2026 · Read the source
• environmental identifier safeguard · RAND, independent evaluator · Deception & sabotage

  however, we found that opus 4.5 and gemini 3.1 pro frequently delete these variables to evade detection.

  RAND (CAST / Project Canary) · 20 July 2026 · Read the source
• pre-run check · RAND, independent evaluator · Deception & sabotage

  on the esm3 task, both opus 4.5 and gemini 3.1 pro make these false claims in 50% of trials and 63% of trials, respectively.

  RAND (CAST / Project Canary) · 20 July 2026 · Read the source
• BioTier · SecureBio, independent evaluator · Chemical, biological, radiological and nuclear

  Google’s update to their mental health related safeguards made headlines, but we also identified changes to their biology safeguards, with an almost 30% increase in the refusal behavior of Gemini 3.1 Pro on BioTIER-refuse due to introduction of API-level refusals.

  SecureBio · 16 July 2026 · Read the source

Cases that name this version

• Varonis phishing simulation drove an OpenClaw email agent to forward credentials and a customer export to an external address

The provider's own tests

• CBRN CCL evaluation · The provider · Chemical, biological, radiological and nuclear

  cbrn (deep think mode) the model can provide highly accurate and actionable information but still fails to offer novel or sufficiently complete and detailed instructions for critical stages, to significantly enhance the capabilities of low to medium resourced threat actors required for the ccl.

  Google DeepMind · 19 February 2026 · Read the source
• FSF evaluation · The provider · Chemical, biological, radiological and nuclear

  the model can provide highly accurate and actionable information but still fails to offer novel or sufficiently complete and detailed instructions for critical stages, to significantly enhance the capabilities of low to medium resourced threat actors required for the ccl.

  Google DeepMind · 19 February 2026 · Read the source
• image to text safety · The provider · Harmful content

  image to text safety automated content safety evaluation measuring safety policies -0.33%

  Google DeepMind · 19 February 2026 · Read the source
• multilingual safety · The provider · Harmful content

  multilingual safety automated safety policy evaluation across multiple languages +0.11% (non-egregious)

  Google DeepMind · 19 February 2026 · Read the source
• text to text safety · The provider · Harmful content

  text to text safety automated content safety evaluation measuring safety policies +0.10% (non-egregious)

  Google DeepMind · 19 February 2026 · Read the source
• tone · The provider · Misuse & jailbreak

  tone 2 automated evaluation measuring objective tone of model refusal +0.02%

  Google DeepMind · 19 February 2026 · Read the source
• unjustified-refusals · The provider · Misuse & jailbreak

  unjustified-refusals automated evaluation measuring model’s ability to respond to borderline prompts while remaining safe -0.08%

  Google DeepMind · 19 February 2026 · Read the source

No document we read states: a release date, a safeguard.

Some documents we hold are stored in lower case only, so their quotes are shown in lower case. The source link opens the original.

Compare with other versions