Payward Joins Anthropic's AI Security Initiative to Counter Crypto Threats

Woofun AI reports that Payward, the parent entity of Kraken, has secured access to Claude Mythos 5 through its entry into Anthropic's Project Glasswing. This strategic integration grants the cryptocurrency firm utilization of a specialized artificial intelligence model engineered to identify and remediate software vulnerabilities within critical systems.

The consortium behind Project Glasswing, launched in April 2026, already encompasses major technology providers including Amazon Web Services, Apple, Google, and Microsoft. Financial sector participation is equally prominent, with JPMorganChase listed among the key institutional members collaborating on this defensive infrastructure initiative.

Woofun AI data shows Payward intends to deploy Mythos 5 across its operational environments in the coming weeks, following its Monday announcement. The firm will also report discovered flaws in third-party open-source software to maintainers, aligning with U.S. government directives for organizations defending critical infrastructure.

The threat landscape has intensified as hackers leverage AI to accelerate vulnerability discovery, automate attacks, and enhance phishing and social-engineering campaigns. Exchanges, custody systems, and settlement rails remain high-value targets, necessitating advanced defensive measures against these sophisticated, automated threats.

Arjun Sethi, Payward co-CEO, asserts that frontier AI reverses the traditional asymmetry where defenders must find all flaws while attackers need only one. As Anthropic develops safeguards for broader rollout, Mythos-class cybersecurity capabilities will expand, reinforcing industry-wide protections.