AI Development

Google, Anthropic and OpenAI unveil cyber-focused AI models and access programs

September 2, 2026 · High importance · model
Editorial Analysis

Google launched Gemini 3.8 Flash Cyber with frontier-level vulnerability detection and automated patching, available to trusted defenders via the Fairwind Program. Anthropic released Claude Fable 5.1 (generally available) and Mythos 5.1 (restricted), while OpenAI confirmed its forthcoming Astra model meets the Critical cybersecurity capability threshold under its Preparedness Framework.

At a Glance
Date
September 2, 2026
Importance
High 4/5
Category
model
Axis of Change
Capability Gain
Organizations
Google Anthropic OpenAI
Models Affected
Gemini 3.8 Flash Cyber Claude Fable 5.1 Claude Mythos 5.1 Astra
Sources
  1. 01 thehackernews.com
Related Developments
Sep 18, 2026 · model

xAI releases Grok Voice Transcribe 2.0 speech-to-text model
Sep 15, 2026 · model

Google releases Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking
Sep 15, 2026 · model

TypeSafe AI emerges from stealth with Jev System One model and $40M seed