TechFlow news, June 19: According to Bloomberg, a select group of early enterprise customers remains able to access the preview version of Anthropic’s AI model “Mythos,” despite a prior U.S. government directive requiring its full shutdown.
According to sources familiar with the matter, multiple institutions—including banks and technology companies—are using Mythos Preview through a testing program codenamed “Project Glasswing.” This initiative comprises approximately 200 authorized organizations, primarily leveraging the AI system for cybersecurity vulnerability detection and research. Cybersecurity firm Dragos confirmed it still has access to the system, and Cisco also verified that it retains continued access rights.
Add to Favorites
Share to Social Media
