Payward, the parent company of crypto exchange Kraken, said on August 17, 2026 that it has joined Anthropic’s Project Glasswing and is putting the Claude Mythos 5 model to work on defensive cybersecurity, marking one of the clearest signs yet that AI security has become an infrastructure priority for major crypto operators.
What Payward Joining Project Glasswing Means
Payward said it is actively incorporating Claude Mythos 5 into its defensive security operations, using the model to scan all Payward environments for vulnerabilities, according to reporting confirming the announcement.
Findings from those scans will be routed into Payward’s existing triage and remediation pipeline rather than a standalone process. The company said any third-party open-source flaws surfaced through the program will be disclosed to maintainers under responsible disclosure practices.
This is not a token launch or a consumer product release. Project Glasswing is Anthropic’s controlled-access program that pairs frontier models with security teams to find and fix software vulnerabilities, and Payward’s participation centers entirely on that defensive use case.
Payward chief executive Arjun Sethi framed the rationale in terms of the asymmetry defenders face.
“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.”
Arjun Sethi, Payward CEO, in the company announcement
Payward said its access follows a U.S. government decision to allow Mythos 5 to reach domestic organizations that operate and defend critical infrastructure. That linkage was asserted by Payward and echoed in secondary coverage, but no standalone government notice has been independently verified, so it should be treated as an unconfirmed claim.
Why AI Security Is Becoming a Bigger Priority in Crypto
A crypto exchange parent joining an external AI security initiative signals institutional concern about safeguarding systems and user funds, where a single unpatched flaw can translate directly into losses. For an exchange-grade operator, feeding automated findings into an existing remediation pipeline is a risk-management step, not a marketing one.
The scale of the program helps explain why participation matters. Anthropic said on June 2, 2026 that it was extending Project Glasswing to approximately 150 new organizations across more than 15 countries, subject to security requirements each partner must meet before gaining access.
The early results give the effort weight. Anthropic said Project Glasswing partners had found more than ten thousand high- or critical-severity vulnerabilities after roughly one month of use, with Cloudflare alone reporting 2,000 bugs, including 400 high- or critical-severity findings.
Anthropic’s initial launch was similarly broad, with the company committing up to $100 million in usage credits plus $4 million in donations and enlisting more than 40 organizations beyond its named founding partners.
Cloudflare, another participant, offered an unusually direct endorsement after pointing the model at more than 50 of its own repositories. Chief security officer Grant Bourzikas said the tooling represents a meaningful improvement over prior approaches.
“Mythos Preview is a real step forward, and it’s worth saying that plainly before getting into anything else.”
Grant Bourzikas, Cloudflare, in a company blog post
What to Watch After the Announcement
The immediate open questions are operational: how quickly Payward’s triage pipeline can act on model-generated findings, and whether the volume of flagged issues strains remediation capacity, as the industry-wide bug counts suggest it might. Future updates from either company will clarify how the deployment performs at exchange scale.
The move also tests whether other crypto firms follow. If a Kraken-scale operator normalizes external AI security partnerships, rivals may face pressure to adopt comparable governance and faster patching standards to keep pace on user trust.
Sentiment across the market remains cautious, with the crypto Fear & Greed Index reading 31, in “Fear” territory. AI-linked tokens have not rallied on the security narrative; benchmark asset Render (RENDER) traded near $1.27, up about 1.5% over 24 hours, against a market capitalization of roughly $659 million, underscoring that this is a security story rather than a speculative one.
Disclaimer: This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency and digital asset markets carry significant risk. Always do your own research before making decisions.
