Research2h ago
Research
OpenAI Releases Astra Cybersecurity Evaluations
Preliminary cybersecurity evaluations for Astra detail the specific safeguards OpenAI is implementing to prevent misuse. The lab focuses on strengthening security controls against critical cyber threats. This transparency provides a baseline for how the company handles model vulnerabilities. Practitioners can now assess the risk profile of these specific capabilities before wider deployment.