According to CoinMeta, OpenAI plans to launch the Astra model soon, but will limit the scope of use for its most advanced cybersecurity features. The company stated on Tuesday that the model will be released “soon,” but did not disclose a specific timeline. The advanced cybersecurity capabilities of Astra were initially available only to a small number of testers, and later access will be provided to more users for defensive purposes through the Daybreak Blue project, along with accompanying cybersecurity protection measures. OpenAI temporarily suspended some internal work in August to enhance security measures. The company said on Tuesday that Astra has reached a “critical level of cybersecurity” and is capable of identifying and developing zero-day vulnerability exploits without human intervention. To prevent abuse, OpenAI has further strengthened the model’s protection measures, including monitoring unauthorized activities and automatically stopping related operations.