Interest is rising in the cybersecurity impact of Anthropic’s AI model Mythos, which is reported to be highly effective at finding software vulnerabilities. As experiences using Mythos are shared by organisations including the Mozilla Foundation and Palo Alto Networks, concerns over AI-driven threats are growing. Security expert Bruce Schneier says such AI could be used not only to attack critical systems but also to exploit loopholes in tax codes and complex regulations, potentially causing greater damage.

U.S. fintech firm Mercury has received conditional approval from the Office of the Comptroller of the Currency for its national bank charter application, moving it to the next stage of launching Mercury Bank. The company said it will keep working with regulators to meet remaining requirements. Mercury is also seeking federal deposit insurance from the FDIC and plans to apply to the Federal Reserve to convert into a bank holding company.