Payward, the parent company of crypto exchange Kraken, has joined Anthropic’s Project Glasswing to use the Claude Mythos model for security research, according to a newly published announcement.

What Payward and Glasswing Announced

The disclosure that Payward is joining Project Glasswing was made in a Payward statement distributed through Business Wire. Payward is the parent company behind the Kraken exchange, which was founded in 2011 and ranks among the longest-running crypto trading platforms still in operation. For related coverage, see Hashdex DEFI Ends Trading Ahead of Closure and Liquidation.

Project Glasswing is the program framework named in the announcement, and Claude Mythos is the Anthropic model referenced for the security-research work. Anthropic, the AI developer founded in 2021 by former OpenAI researchers and best known for the Claude model family, has published a Claude Mythos preview system card describing the model.

Payward also flagged the news through its official account on X: https://x.com/Payward/status/2089428414675439983. Verification of the arrangement is currently only partial, and specific deployment details, performance figures, or customer outcomes were not confirmed in the available material. Those omissions also define what to watch next: whether the scope of the deployment is described, whether any evaluation or performance data is published, and whether additional confirmation follows beyond the distributed statement.

Payward has been in the news on other fronts recently, including reports that the company cut roughly 150 jobs ahead of a planned IPO and moves to expand its xStocks product into Hong Kong, the UK, and South Korea.

Why Claude Mythos Security Research Matters for Crypto Firms

For a crypto operator, the relevance of an AI security-research model sits in operations rather than token prices. Security tooling touches exchange infrastructure, custody stacks, and blockchain threat analysis, all areas where an exchange parent like Payward carries real exposure. That kind of exposure has precedent: exchange security has been an existential industry issue since well-documented episodes such as the 2014 Mt. Gox collapse and the 2016 Bitfinex theft. Kraken already operates an in-house security research arm, Kraken Security Labs, which has published findings on hardware wallets and other crypto infrastructure.

Potential applications for a model positioned in the Claude Mythos preview include smart-contract review, vulnerability triage, and threat-intelligence analysis. These remain possible use cases, not confirmed Payward deployments, given the partial verification of the announcement.

The broader pattern of AI and blockchain infrastructure convergence has been visible elsewhere in the market, from large AI cloud deployments to stablecoin issuers exploring AI applications. Specialized AI security tooling, if it delivers on that framing, could shape how crypto firms run risk operations over time.

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.