Kraken’s parent company Payward has joined Anthropic’s Project Glasswing and is using the AI model Claude Mythos 5 to hunt for software vulnerabilities across its systems — a move that puts one of crypto’s biggest players at the forefront of AI-assisted cybersecurity. The decision follows a coordinated appeal last week from more than 40 Bitcoin and crypto firms — including Kraken, Coinbase, Block and BitGo — urging Anthropic, OpenAI and other labs to give vetted defenders access to their most powerful models. Organized by the Bitcoin Policy Institute, the letter argued that teams protecting open-source financial infrastructure need frontier AI tools to find and fix flaws before malicious actors can exploit them. Payward said on Monday it will run Mythos 5 scans against its codebase and infrastructure, with any flagged issues routed to its security teams for triage. It is the first reported crypto company to join Project Glasswing and gain access to Claude Mythos 5. “Selection gives Payward’s security division early access to the same class of model, sharpening its ability to combat sophisticated software vulnerabilities and protect millions of customers across the globe,” the company wrote. Payward also said it will responsibly disclose bugs it finds in third‑party open-source software to project maintainers. The company framed the approach as a way to both harden its own financial infrastructure and improve open-source tooling that underpins wide swaths of the crypto industry. Project Glasswing is Anthropic’s vetted cybersecurity program that grants qualified organizations access to its most capable cyber models. Launched in April and expanded in June, Glasswing partners have reportedly uncovered thousands of high- or critical-severity vulnerabilities. Anthropic did not immediately respond to requests for comment. The move comes amid growing evidence that advanced AI can be a double-edged sword for software security. In April, Mozilla said Anthropic’s Claude Mythos flagged 271 vulnerabilities in Firefox during internal testing — an example of how the same models that help defenders can also be used to generate more effective exploits. Payward Co-CEO Arjun Sethi framed the shift as a practical response to a long-standing defensive disadvantage: “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.” Why it matters for crypto: by gaining early access to frontier cyber models, crypto firms can better protect complex, open-source financial infrastructure. But the trend also underscores the urgency behind calls for responsible, vetted access to powerful AI tools so defenders—not attackers—gain the upper hand. Read more AI-generated news on: undefined/news