Hitachi Uses Anthropic's AI 'Claude Muitos' for Product Security Verification, Cutting Attack Scenario
Hitachi announced it used Anthropic's latest AI, 'Claude Muitos,' to verify threats to its products and software. The verification reduced attack scenario creation time from weeks to hours, improved vulnerability detection accuracy, and significantly eased developer burden. It covered more than 100 social infrastructure cases in fields including energy. Hitachi plans to implement product updates based on the results. In June, it became the first Japanese manufacturer to gain access to Muitos, which is strictly controlled to prevent misuse.
Hitachi announced on the 28th that it used the latest artificial intelligence 'Claude Muitos' from US startup Anthropic to verify threats to its own products and software. The verification showed that the time required to create attack scenarios was reduced from several weeks to several hours, the accuracy of system vulnerability detection improved, and the burden on developers was significantly reduced.
The verification covered more than 100 social infrastructure cases in fields including energy. The verification found that the AI can perform large-scale software analysis at high speed, but final evaluation and risk judgment still require human experts. Hitachi will implement product updates sequentially based on the verification results. In June, Hitachi became the first manufacturer in Japan to gain access to 'Muitos.' Muitos is particularly capable of discovering security vulnerabilities, and access to it is strictly restricted to prevent misuse.