# Cisco and IBM Plan Separate Claude Mythos Security Integrations By Simon Yoon Canonical URL: https://www.tokenpost.com/news/technology/53465 Published: 2026-11-11T01:41:37.000Z Updated: 2026-11-11T01:41:37.000Z Section: Technology > Cisco’s AI Defense targets AI testing, while IBM plans code scans and exploitability checks through Concert Protect, with general availability aimed for early 2027. Cisco and IBM plan to add Anthropic’s Claude Mythos to separate security products, with Cisco targeting AI systems and IBM focusing on weaknesses in customer code for enterprise defense. Cisco’s AI Defense product will use Mythos to help plan attacks that account for a system’s context and unfold across multiple steps. Access will run through a structured interface without open-ended prompts. Cisco’s controls limit tests to registered, authorized targets and named operators; sessions expire, can be halted and leave audit records. The feature is coming soon to select enterprises, but Cisco has not given a firm release date. “Enterprises are increasingly building AI agents into their core operations and want to test those agents thoroughly before putting them into production,” said Anthropic Cybersecurity Lead Michael Moore. IBM plans to use Mythos in Concert Protect to scan code repositories selected by customers and help assess whether weaknesses can be exploited. The integration is also designed to generate proposed fixes for developers to review and approve. IBM is targeting general availability in early 2027. IBM’s earlier work with Mythos focused on internal security research and evaluation. The planned Concert Protect integration is a step toward bringing that work into enterprise customer workflows. Anthropic announced its [Project Glasswing security program]() in April, giving a limited group of partners access to Mythos for defensive security work. Cisco was among the launch partners. Anthropic continues to limit access to a small set of vetted organizations; it describes Mythos as capable in cybersecurity and biology research, with potential for beneficial and harmful use. Neither company has reported detection rates or customer results for the planned integrations. ## Links in this article - [Project Glasswing security program](https://www.tokenpost.com/news/technology/33957)