Rubrik has been granted access to Anthropic’s Mythos Research Preview as part of Project Glasswing — an initiative that hands frontier AI to companies maintaining critical software so they can find and fix vulnerabilities before attackers do.
The security and AI operations company says it is deploying Mythos defensively: identifying, vetting, and patching potential weaknesses across its enterprise platform and product suites, proactively purging vulnerabilities before they can ever be leveraged against it.
Project Glasswing is Anthropic’s answer to an uncomfortable shift in cybersecurity: AI has dramatically accelerated vulnerability discovery, and the same capability that helps defenders harden systems can help attackers find ways in. The project’s stated goal is “to use Mythos Preview to help secure the world’s most critical software, and to prepare the industry for the practices we all will need to adopt to keep ahead of cyberattackers.”
Access to the Mythos preview is restricted to organizations that build or maintain critical software infrastructure. As a member of the Glasswing cohort, Rubrik commits to sharing findings back to the industry — extending the benefit beyond its own codebase.
For Rubrik, whose business is built on cyber resilience — keeping enterprise data recoverable when attacks succeed — the move closes the loop: using AI to prevent breaches in the very platform customers rely on after one.
The announcement is also a signal of where security practice is heading. Vulnerability research is becoming an AI-versus-AI race, and vendors of critical infrastructure increasingly cannot afford to run it with humans alone.
