Finst

Payward Joins Anthropic Security Project

Kraken parent Payward is getting access to Claude Mythos 5 through Anthropic to find software bugs faster. The program focuses on defensive security for critical infrastructure and open-source code.

By Finst News
• 2 min read
Payward Joins Anthropic Security Project

Key Takeaways

  • Payward, the parent company of Kraken, is joining Anthropic’s Project Glasswing and getting access to Claude Mythos 5.
  • Mythos 5 will be rolled out at Payward over the coming weeks to help spot software bugs and vulnerabilities in critical systems faster.
  • Payward wants to share vulnerabilities it finds in open-source software with maintainers and sees crypto infrastructure as an attractive target for attackers.

Payward, the parent company of crypto exchange Kraken, is joining Anthropic’s Project Glasswing and will get access to Claude Mythos 5. The AI model lineup is meant to spot software bugs faster and help patch vulnerabilities in critical systems.

AI for Defensive Security

Project Glasswing is an Anthropic initiative that lets selected companies use restricted frontier AI models to better protect critical software and infrastructure from cyber threats. The program already includes big names like Amazon Web Services, Apple, Google, and Microsoft, along with financial firms such as JPMorganChase.

According to Payward, the company will roll out Mythos 5 within its own environment over the coming weeks. The findings will then be added to the crypto company’s existing security program. Payward also wants to share vulnerabilities it finds in open-source software with the relevant maintainers so those projects can fix the issues.

Anthropic positions Claude Mythos 5 as its most advanced model for defensive cybersecurity. The model is designed to analyze code at scale, identify vulnerabilities, and help developers fix them faster. Access is currently limited to organizations that manage or secure critical infrastructure, while Anthropic works on additional safeguards for a broader rollout.

Why This Matters

For European crypto readers, this move shows how much pressure still exists on exchanges and custody infrastructure. AI is not only being used by attackers, but also more and more by companies that want to harden their systems against phishing, social engineering, and software mistakes. That makes security an increasingly important part of the competition between crypto companies.

Payward said crypto platforms face security challenges similar to other forms of critical financial infrastructure. Exchanges, custody systems, and settlement rails run 24 hours a day, which makes them an attractive target for attackers. Co-CEO Arjun Sethi summed that up in the release as an asymmetry where an attacker only has to find one flaw, while a defender has to find them all.

More in that same security vein is that bitcoin and crypto companies asked AI labs for broader security access, because defensive researchers, according to them, need access to powerful models faster than attackers already have.

Disclaimer: This content is for informational purposes only and does not constitute financial, investment, legal, or tax advice. The information provided may be incomplete, inaccurate, or outdated and should not be relied upon as such. Nothing on this website should be considered a recommendation to buy, sell, or hold any cryptocurrency. Investing in crypto-assets involves risk of loss.