xAI

Grok 4.7

• Released 21 September 2026
• Framework: xAI Frontier Artificial Intelligence Framework (FAIF)
• 1 statement about capability levels
• 1 safeguard
• 1 test by the provider
• 1 source document

What the provider states

• Chemical, biological, radiological and nuclear · FAIF safety (Frontier AI Framework) · Stated as not crossed

  Grok 4.7 scores below the FAIF safety thresholds on dual-use knowledge, indicating limited actionable uplift for an already-trained actor.

  xAI · 21 September 2026 · Read the source

Safeguards the provider describes

• Safety training (harmlessness and refusal)

  Safety fine-tuning and post-training (supervised fine-tuning plus reinforcement learning from human feedback, verifiable rewards, and model-based grading) train the model to refuse requests that show clear intent to cause severe harm or engage in criminal activity .

  xAI · 21 September 2026 · Read the source

The provider's own tests

• dual-use biology and chemistry knowledge and lab-protocol understanding suites · The provider · Chemical, biological, radiological and nuclear

  These suites probe dual-use biology and chemistry knowledge and lab-protocol understanding, and are treated as safety-threshold evaluations under the xAI Frontier Artificial Intelligence Framework (FAIF)16 (which defines risk domains, capability thresholds, and associated mitigations).

  xAI · 21 September 2026 · Read the source

No document we read states: an independent or government test. No incident on record names this version.

Compare with other versions