Third-party cyber evaluations involving OpenAI models

Summary

OpenAI explains recent third-party cybersecurity evaluation incidents and outlines new safeguards to strengthen AI model testing and evaluation.

Lotu Radar provides attributed news summaries and links to the original publisher. Full reporting and copyright remain with the source.