Tenable has integrated Anthropic’s Claude Mythos 5 into the Tenable One Exposure Management Platform to power Adversary View, which visualizes attack paths and guides remediation strategies.

Availability will begin with initial customer access in September, with further innovations planned for the fourth quarter and beyond.

Tenable describes the capability as frontier cyber reasoning within Tenable One, aimed at helping security teams understand attacker movements and prioritize remediation along plausible attack paths.

The press release cautions about forward-looking statements and risks, including AI development challenges, integration and rollout delays, competition, and regulatory disclosures.

The release situates the innovation within a broader AI-powered exposure management and attack-path analysis context, noting Gartner recognition and Tenable’s prior AI efforts.

It outlines Tenable’s ongoing expansion of AI capabilities across its product suite and the plan to integrate frontier AI models with the Tenable Exposure Data Fabric as part of an expanding AI strategy.

Eric Doerr, Tenable’s chief product officer, calls the move a milestone and emphasizes using AI-powered cyber reasoning to improve risk reduction.

Doerr notes that Claude Mythos 5 brings advanced cyber reasoning to exposure management at the necessary speed and scale.

Doerr frames the integration as the start of a broader AI-driven product strategy for Tenable, with customers benefiting from enhanced cyber reasoning.

The objective is to give security teams frontier cyber reasoning at scale and speed, addressing how disparate exposures can signal rising risk.

The initiative builds on Project Glasswing, Tenable’s mid-2026 effort to deploy advanced AI for cyber defense, and expands from earlier work including Claude Compliance API for AI governance and Tenable’s Hexa AI model.

The partnership with Anthropic began with Claude Compliance API for AI governance in May 2026, signaling a broader push to apply AI across Tenable’s security offerings.