Hitachi announced on the 28th that it has significantly streamlined the verification of cybersecurity threats against its products and software by utilizing “Claude Mythos,” the latest high-performance AI developed by U.S. AI startup Anthropic.
According to the announcement, a series of safety verification tasks—including the creation of scenarios for unauthorized access and cyberattacks, as well as the detection of system vulnerabilities—that previously required several weeks can now be completed in just a few hours. This has substantially reduced the burden on developers.
For this initiative, Hitachi used the preview version of Mythos, “Mythos Preview.” The company applied it to over 100 cases involving products and systems for social infrastructure, primarily in the energy sector, to verify its effectiveness.
In the specific verification work, the AI rapidly analyzed massive source code bases comprising millions of lines. It succeeded in identifying latent vulnerabilities within the systems with high precision.
On the other hand, Hitachi explained that while the series of verifications demonstrated the AI’s ability to execute large-scale software analysis at high speed, it also became clear that the involvement of human experts is indispensable for final risk assessments and security judgments.
Based on these verification results, Hitachi plans to sequentially implement security updates for the targeted products.
This announcement is drawing attention as a case where a major company handling critical infrastructure has begun the full-scale deployment of generative AI in the field of actual product security enhancement. It is particularly significant that the time-reduction effects and accuracy improvements achieved by AI in highly specialized areas such as code analysis and threat modeling have been demonstrated with concrete metrics.
Anthropic’s Claude series has a strong reputation as a large language model (LLM) offering advanced reasoning capabilities and long-context processing for enterprises. Hitachi’s case demonstrates that these capabilities also prove highly effective in practical domains that underpin the robustness of social infrastructure.
Add to Google Preferred Sources
Once added, BigGo Finance appears first in Google Search Top Stories, so you get the broadest, most up-to-the-minute, and most comprehensive global financial news first.