Kraken Parent Payward Joins Anthropic's Project Glasswing to Bolster Crypto Security

Payward, the parent company of cryptocurrency exchange Kraken, has joined Anthropic's Project Glasswing, a security-focused initiative that gives participating organizations access to Claude Mythos, a specialized AI model designed to identify and analyze cybersecurity vulnerabilities.

The move highlights the growing overlap between artificial intelligence and blockchain security, as crypto companies increasingly turn to advanced AI tools to help protect user assets and platform infrastructure from sophisticated threats.

What Project Glasswing Is

Project Glasswing is an Anthropic initiative that brings together select organizations to use AI capabilities for security research and vulnerability detection. Members of the program receive access to Claude Mythos, a version of Anthropic's Claude AI designed for security-related tasks such as identifying weaknesses in code, systems, and protocols.

By joining the program, Payward plans to use Claude Mythos to proactively search for security flaws across its platforms before malicious actors can exploit them. This type of AI-assisted security research marks a significant development from traditional manual penetration testing and bug bounty programs, with the potential to make vulnerability discovery faster and more comprehensive.

As reported by Decrypt, Payward's participation in Project Glasswing reflects the company's commitment to deploying advanced technology in its security operations, a priority for any exchange that handles billions of dollars in digital assets.

Why It Matters for the Crypto Industry

Security remains one of the most pressing issues in the cryptocurrency sector. Exchanges and custodial platforms are frequent targets for hackers, and even small vulnerabilities can lead to significant financial losses for users and institutions.

Integrating AI-powered security tools into a major exchange's defense strategy could establish a new industry benchmark. Kraken is one of the longest-running and most widely used cryptocurrency exchanges globally, and Payward's decision to participate in a structured AI security program — rather than relying only on internal teams or third-party auditors — shows how seriously the company is treating the evolving threat landscape.

The use of a purpose-built AI model such as Claude Mythos also points to a broader trend: AI systems are no longer being used only for customer-facing applications or trading analytics, but are increasingly being applied to infrastructure protection and risk mitigation in financial technology. For firms operating large digital asset platforms, that shift matters because security has become part of the product experience itself, shaping user trust and operational resilience alongside trading access and liquidity.

Anthropic's Role in Crypto Security

Anthropic, the AI safety company behind the Claude family of models, has been expanding its enterprise partnerships across multiple industries. Project Glasswing appears to be one of its more specialized programs, aimed at organizations where security vulnerabilities can have outsized consequences.

By providing access to Claude Mythos through this initiative, Anthropic is positioning itself as a key partner for industries — including crypto — where the cost of a security breach can be catastrophic. The program also gives Anthropic real-world feedback on how its AI models perform in high-stakes security environments, which could help shape future model development.

Implications for Kraken Users and the Broader Market

For Kraken users, the partnership offers some reassurance that the exchange is investing in next-generation security measures. While no security system is completely foolproof, the proactive use of AI to identify flaws before they can be exploited adds another layer of defense.

More broadly, Payward's move could encourage other major exchanges and crypto custodians to explore similar AI-driven security programs. As the industry matures and regulatory scrutiny increases, demonstrating strong security practices is becoming as important as liquidity and trading features in attracting institutional clients. That makes programs focused on vulnerability detection especially relevant, since they address a risk area that can affect both day-to-day users and larger market participants.

The partnership between Payward and Anthropic through Project Glasswing reflects a broader recognition that the tools used to secure digital asset platforms must evolve alongside the threats they face — and that AI is becoming increasingly central to that effort.