Anthropic Expands Access To Mythos 5 Cybersecurity Capabilities / Fresh Today / CUToday.info | #hacking | #cybersecurity | #infosec | #comptia | #pentest | #ransomware

SAN FRANCISCO—Anthropic is expanding access to the cybersecurity capabilities of its advanced AI models, including Claude Mythos 5, while using controls designed to keep the technology from being misused by attackers, according to SecurityWeek.

The effort includes integrations with cybersecurity vendors, an upgraded Claude Security offering, a $35-million open-source security fund and plans to broaden Anthropic’s Cyber Verification Program.

SecurityWeek reported the expansion builds on Project Glasswing, launched in April to give a limited group of organizations early access to Claude Mythos Preview and later Mythos 5. Anthropic is now working with cybersecurity companies to incorporate Mythos 5 into security operations, incident response and detection tools used to protect critical infrastructure and other organizations. Rather than interacting directly with the model, users will receive narrowly defined defensive outputs, such as recommended patches or security alerts.

Claude Security, now in public beta for Claude Enterprise customers, is also using Mythos 5 to scan codebases, SecurityWeek reported. The service provides findings that include CWE categories, confidence and severity ratings and suggested fixes, while requiring human approval before fixes are deployed.

“Claude Security uses Mythos 5 to scan code you own, and returns detailed findings rather than raw outputs without exposing the model itself,” Anthropic said. “This means defenders can access the capabilities of Claude Mythos 5 without the model becoming accessible to those who might misuse it.”

SecurityWeek also reported Anthropic has created the Defender Advantage Fund, or 0xDAF, committing $35 million in Claude credits to organizations helping open-source maintainers improve security. Anthropic also plans to expand its Cyber Verification Program in coming weeks to allow vetted organizations broader dual-use capabilities for authorized security work, including vulnerability triage and validation, with Mythos-class access expected later. The company is separately continuing to expand Mythos access through Project Glasswing with U.S. government partners, particularly for organizations protecting critical infrastructure.