Search
On September 2, it was reported that OpenAI plans to launch the Astra model soon, but will restrict the use of its most advanced cybersecurity features. The company stated on Tuesday that the model will be released 'soon,' without specifying an exact date. Initially, Astra's advanced cybersecurity capabilities will only be available to a limited number of testers. Subsequently, through the Daybreak Blue project, more users will be granted defensive access with cybersecurity protections in place. Previously, in August, OpenAI had paused some internal work on Astra to enhance security. Currently, Astra has reached a 'critical cybersecurity threshold,' enabling it to identify and develop zero-day exploits without human intervention. To prevent misuse, OpenAI has strengthened model protections, including monitoring and automatically stopping unauthorized activities.
-
C114 Communication Network -
Communication Home
