[Anthropic Opens Mythos User Network for Sharing Cyber Threat Intelligence] According to a report by Jinse Finance, on May 19, out of concern that restricted access to information could harm the interests of small and medium-sized enterprises, Anthropic recently adjusted its original stance and officially allowed users of its high-performance AI model, Mythos, to share cyber threat intelligence with other organizations facing similar security vulnerabilities. This new policy highlights the practical challenges leading AI companies face in strictly controlling access to top-tier models. Previously, Anthropic, through a project called the Glasswing Initiative, only granted Mythos access to approximately 50 large enterprises and institutions responsible for core digital infrastructure. The model's efficiency in identifying software vulnerabilities far surpasses manual inspections. The project's initial goal was to enable these enterprises and government institutions to address cybersecurity risks in advance, then gradually expand access and introduce more powerful AI models. An Anthropic spokesperson stated that the initial users of Mythos had signed confidentiality agreements, strictly prohibiting the disclosure of cybersecurity risk information they possessed. However, last week, Anthropic formally notified its partner organizations that, under the principles of compliance and prudence, they could share cyber threat intelligence and the security vulnerabilities identified by the Mythos model with other institutions.