OpenAI reveals preliminary cyber evaluations of Astra and announces stronger safeguards
OpenAI
OpenAI has published initial cybersecurity assessments for its upcoming AI model Astra, detailing the evaluation results and outlining additional security measures being implemented to mitigate potential risks. The company emphasizes its commitment to responsible development and deployment of advanced AI capabilities.
OpenAI has released preliminary cybersecurity evaluations for Astra, a new AI model under development, as part of its ongoing efforts to address critical cyber capabilities. The evaluations outline potential risks and vulnerabilities associated with Astra, and OpenAI is taking proactive steps to strengthen safeguards and security controls to mitigate these risks. The company remains focused on ensuring that its AI technologies are developed and deployed responsibly, with robust security measures in place.
Source: OpenAI —
original