Crypto teams join queue for Anthropic's AI vulnerability scanner
Anthropic's opt-in service uses its most capable models, including Claude Mythos, to produce vulnerability reports, and crypto projects are applying.
Anthropic has introduced an opt-in service that generates vulnerability reports using its most capable AI models, including a model it calls Claude Mythos, according to a Cointelegraph report. Crypto projects are among the applicants.
The offering is positioned as an AI-driven complement to conventional code review: teams submit their work, and the models return a report on potential weaknesses. The source did not name the crypto projects involved, disclose pricing, or specify the scope of a typical scan.
Why protocol teams are watching
Security review is a standard line item for protocols before and after a launch, and smart-contract audits can be slow and costly to schedule. An AI-generated report could serve as a first pass ahead of, or alongside, a human audit, though the reliability of automated findings remains an open question for many engineering teams.
AI vendors and crypto firms have been moving toward each other on security tooling over the past year, as exploits and bridge failures keep drawing attention to code-level risk. Whether model-generated vulnerability reports become a routine part of crypto development pipelines will depend on how their findings hold up under human scrutiny.