Visa (V) developed in April a harness that reduces the cost of and time it takes for Anthropic's Mythos AI to find and fix cyber security vulnerabilities in its codebases significantly, reports The Information

Internal tooling stories of this kind, a large company building bespoke harnesses around a frontier model for security work, have become a recurring genre as enterprises move from piloting AI to operationalising it, and they have tended to matter more as signals of adoption depth than as standalone catalysts. The relevant read here is twofold: for Visa it speaks to the cost trajectory of maintaining large legacy codebases, where automated vulnerability remediation compresses a labour-intensive line item, though the equity rarely re-rates on engineering efficiency alone; for the model provider it is another data point in the enterprise penetration pattern that has underpinned the private-market narrative around frontier labs. The case distinction worth drawing is between internal productivity use cases, which flow quietly through opex, and productised AI features, which carry revenue implications, and this is the former. Worth watching is whether similar deployments surface at peer financials, which would mark the shift from anecdote to sector practice, and any disclosure in subsequent results on technology spend or headcount mix. As a single-source press report of an internal tool, the headline is incremental rather than market-moving.