Google—Google withheld cyber-capable Gemini 4 Argon from UK's AI Security Institute for independent testing, prioritizing US government and its own Wiz subsidiary instead
Ahead of the limited September 30, 2026 release of Gemini 4 Argon, Google did not give the UK government's AI Security Institute (AISI) pre-release access to the version without cyber safety guardrails for independent safety testing, departing from the access AISI had previously received for prior frontier models. Instead, early access went to the US government, at the Trump administration's request, and to Wiz, the Israeli cybersecurity firm Google owns, as part of a limited 650-member 'Fairwind Program' rollout. OpenAI and Anthropic reportedly made similar access decisions favoring the US government over AISI around the same period. Google said it would give AISI access 'before broader release.'
Scoring Impact
| Topic | Direction | Relevance | Contribution |
|---|---|---|---|
| AI Oversight | -against | primary | -1.00 |
| Overall incident score = | -0.572 | ||
Score = avg(topic contributions) × significance (medium ×1) × confidence (0.57)
Evidence (1 signal)
Politico Europe: Google bypasses UK AI safety testing for initial release of Gemini 4 Argon
Politico reported that Google did not provide the UK AI Security Institute with pre-release access to the unrestricted version of Gemini 4 Argon, instead prioritizing the US government and its own Wiz subsidiary, departing from AISI's previous privileged access to frontier Google models.