U.S. AI developer Anthropic announced on October 8 the launch of “Cyber Mission,” a new program aimed at protecting critical infrastructure—including power grids, water systems, manufacturing, and transportation—from cyberattacks. Eleven companies worldwide have joined as initial members, including Hitachi (6501.T) and Nozomi Networks, a U.S.-based subsidiary of Mitsubishi Electric.
As AI-powered cyberattacks grow increasingly sophisticated, the coalition will pool operational technology and security expertise from each member to assess and remediate vulnerabilities in the control systems that underpin social infrastructure. The goal is to protect OT (operational technology) networks connecting critical infrastructure with factory equipment, preventing system failures and production stoppages before they occur.
The program’s core component, the “Critical Infrastructure Defense Program,” targets service providers that support operators responsible for running power, water, manufacturing, and transportation networks. It combines AI tools, engineering support, and funding to strengthen the defensive capabilities of those on the front lines.
In addition to Hitachi and Nozomi Networks, founding members include consulting giant Accenture and U.S. security leaders CrowdStrike and Palo Alto Networks. Hitachi will participate in working groups and share expertise in equipment control and operational technology cultivated through its work in critical infrastructure sectors.
CrowdStrike has disclosed that it is partnering with Anthropic by combining its own security tools with Anthropic’s AI models in an integrated offering.
Anthropic simultaneously introduced “OSS Scanner,” a free service for open-source projects. The system uses AI to automatically generate vulnerability reports and deliver them to project maintainers, with participation being voluntary. However, the company acknowledged that these reports have not undergone human review and may contain inaccuracies.
Concerns have emerged from the open-source community. Google has pointed out that automated security findings could increase the triage burden on maintainers. Discussions within the Open Source Security Foundation (OpenSSF) have also warned that AI-generated disclosures risk adding noise to the ecosystem.
Earlier this week, Anthropic integrated a project that expands defensive access to its AI models into its cyber evaluation program, accelerating its push to apply AI in the critical infrastructure space.
The broader security industry is increasingly marketing AI-powered cyber defense tools to critical infrastructure operators. Anthropic’s latest initiative exemplifies this growing trend.