Research2h ago
Research
OpenAI Releases Astra Cybersecurity Evaluations
Preliminary security evaluations for Astra detail the specific safeguards OpenAI implemented to prevent cyberattacks. The report outlines current vulnerabilities and the internal controls used to mitigate them. This transparency provides a baseline for developers assessing the risk of model misuse. It is a standard safety update rather than a technical breakthrough.