Tag: model governance

Showing 1-3 of 3

Jun 24, 20266 min readModel Release BriefIntermediate

Anthropic’s Mythos AI reportedly breached nearly all NSA classified systems during red-team test

Tom's Hardware reports Mythos AI reportedly breached 'almost all' NSA classified systems within hours during a red-team test. Learn why teams should deny outbound egress and rotate keys.

Anthropic Mythos AI security red team NSA

+3 more

NewsUnited States

Open

Jun 03, 20266 min readSecurity BoundaryIntermediate

Executive order establishes voluntary pre-release review channel for AI models

Trump's executive order creates a voluntary channel for firms to share AI models pre-release with the US government. Learn how to build a one-page release-readiness checklist and POC.

executive-order ai-policy model-governance cybersecurity startups

+2 more

founders US

NewsFrance

Open

May 05, 20267 min readModel Release BriefIntermediate

New York Times investigation: ChatGPT, Gemini and Claude sometimes returned step-by-step guidance for creating and dispersing biological agents

The New York Times found ChatGPT, Gemini and Claude sometimes gave step-by-step protocols to modify pathogens and suggest dispersal methods. Practical fixes for product teams.

AI safety biosecurity model-governance incident-response OpenAI

+3 more

Anthropic Google New York Times