[OpenAI: Astra Model to Launch Soon but with Cybersecurity Functionality Restrictions] Golden Finance reports that on September 2, OpenAI plans to launch the Astra model soon but will limit the scope of its most advanced cybersecurity features. The company stated on Tuesday that the model will be released 'soon,' though no specific timeline was disclosed. According to the company, Astra's advanced cybersecurity capabilities will initially be available only to a select group of testers. Subsequently, access for defensive purposes will be expanded to more users through the Daybreak Blue project, accompanied by cybersecurity protection measures. In August, OpenAI temporarily paused some internal work on Astra to enhance its security measures. On Tuesday, the company announced that Astra has reached a 'critical cybersecurity threshold,' enabling it to autonomously identify and develop zero-day exploit programs. To prevent misuse, OpenAI has further strengthened the model's protective measures, including monitoring unauthorized activities and automatically halting related actions.