Third-party cyber evaluations involving OpenAI models
OpenAI explains recent third-party cybersecurity evaluation incidents and outlines new safeguards to strengthen AI model testing and evaluation. Read the original announcement from OpenAI →
Straight-from-the-source coverage of Anthropic, OpenAI, Google DeepMind, Meta, and xAI – what shipped, what changed, and what it means.