OpenAI shares cybersecurity evaluations and steps to strengthen Astra safeguards

OpenAI has disclosed preliminary cybersecurity evaluations for its Astra model along with measures to strengthen safeguards and security controls, addressing concerns about the system's capabilities.