Google Launches Gemini 4 Argon With Guardrail-Free Access for Vetted Defenders
Google launches frontier model Gemini 4 Argon, initially without cyber guardrails for vetted defenders.
Google announced Gemini 4 Argon, a frontier model initially offered to trusted cyber defenders and internal teams through the Fairwind Program without cyber guardrails. Google says Argon can autonomously find, validate, and patch vulnerabilities and uncovered an unnamed critical flaw exposing personal data in hospital software used worldwide. On Collinear AI's CWE-bench v1, Argon tied for first at 68% with OpenAI's GPT-6 Astra and xAI's Grok 4.7. A broader release is intended to refuse cyber and CBRN misuse, monitor chain-of-thought for misalignment, and resist indirect prompt injection.
- Limited Fairwind access removes cyber guardrails for vetted defenders.
- Argon reportedly found a critical vulnerability in unnamed hospital software.
- CWE-bench v1 score of 68% ties GPT-6 Astra and Grok 4.7.
- Wider release adds CBRN refusals and prompt-injection mitigations.
- Vendors
- Products
- CodeMender
- Organizations
- GoogleWizOpenAIxAICollinear AI
- Countries
- United States
From the original article
3 min read at securityweek.com
The company says its new frontier AI model found a critical vulnerability in software used by hospitals worldwide. The post Google Launches Gemini 4 Argon With Guardrail-Free Access for Vetted Defenders appeared first on SecurityWeek.
By Eduard Kovacs · securityweek.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.