Payward Taps Anthropic’s Claude Mythos 5 to Fortify Global Crypto Infrastructure

Payward, Inc., the parent company of Kraken, has integrated Anthropic’s Claude Mythos 5 model into its cybersecurity operations via Project Glasswing. For fintech professionals, this represents a critical shift in defensive strategy, utilizing frontier AI to automate the identification of software vulnerabilities at a scale that human red teams cannot match.

What was announced

Payward has been selected as a participant in Project Glasswing, an initiative launched by Anthropic in April 2026 to secure critical global infrastructure. The partnership grants Payward’s security division early access to Claude Mythos 5, a high-tier AI model capable of identifying and exploiting software vulnerabilities with a proficiency that rivals or exceeds highly skilled human actors. This access follows a US government decision to permit the deployment of Mythos 5 for organizations responsible for defending critical infrastructure.

The integration is designed to bolster Payward’s existing defensive posture. The company will use Mythos 5 to scan all internal environments for security flaws. Once identified, these vulnerabilities are fed into Payward’s established triage and remediation pipeline, where they are addressed alongside findings from the company’s internal red and blue teams and its long-standing bug bounty program. Since Project Glasswing's inception, partners have identified thousands of high or critical-severity flaws, highlighting the model's efficacy in pre-empting cyberattacks.

Furthermore, Payward has committed to an open-source contribution model. When Mythos 5 identifies vulnerabilities within third-party open-source software that Payward utilizes, the company intends to report these findings upstream to the relevant project maintainers. This ensures that security patches are distributed throughout the broader crypto and financial ecosystems, rather than remaining siloed within a single firm.

"Security has always been an unfair game. An attacker needs to find one flaw. A defender has to find all of them, first, every single day. Frontier AI is the first thing that flips that asymmetry. A model can read every line of code the way an attacker would, at machine scale, so we find the flaw before anyone can build the exploit,"

Arjun Sethi at Payward.

The companies involved

Payward, Inc. operates as a unified financial infrastructure platform with a mission to advance an open global financial system. It is most widely recognized as the parent company of Kraken, one of the world’s most prominent cryptocurrency exchanges. Payward has maintained a rigorous focus on security for over a decade, holding industry-standard certifications including ISO 27001 and SOC 2. The company’s internal security culture is defined by constant "war-gaming," where dedicated red teams simulate real-world attacks to test the resilience of the blue teams’ defensive hardening.

Anthropic is the AI safety and research company behind the Claude series of models. By launching Project Glasswing, Anthropic has positioned itself as a central player in the defense of critical infrastructure, partnering with leaders in technology and finance to ensure that the same AI capabilities that could be used for offensive cyber warfare are first harnessed for defensive purposes. The collaboration with Payward marks a significant expansion of Anthropic's influence within the regulated financial services and digital asset sectors.

What this means

This move signals the end of the era where manual code review and periodic penetration testing were sufficient for tier-one financial institutions. By deploying Mythos 5, Payward is acknowledging that the speed of AI-driven offense requires an equally automated defense. The "asymmetry" Arjun Sethi references is a genuine pain point for CISOs; this partnership suggests that the advantage may finally be shifting toward defenders who can afford to run frontier models at scale.

The industry should watch how other major exchanges respond. If Payward successfully hardens its infrastructure and contributes to open-source security, it sets a new benchmark for "responsible disclosure" in the crypto space. Competitors now face pressure to prove their defensive stacks are equally sophisticated, or risk being viewed as the path of least resistance for AI-powered attackers.