Anthropic unveils the 'Cyber Mission' and a free AI scanner for open-source software
Rather than just chasing raw capability, Anthropic is deploying engineering talent, tools, and funding to help secure the open-source and critical-infrastructure systems that run the modern world.
Anthropic has launched a new 'Cyber Mission', a long-term commitment to securing the systems that everyone relies on, anchored by a tool called OSS Scanner that offers open-source projects free, regular security scans from the company's strongest AI models.
What the Cyber Mission actually is
Announced on October 8, the Cyber Mission is framed not as a single product but as an ongoing program. According to the announcement, Anthropic will deploy its own engineering talent and provide tools, research, and funding to the people and organizations that secure critical infrastructure and open-source software, with the goal of releasing tools, research, and resources in new areas over time.
At the center is OSS Scanner, which Anthropic describes as offering open-source projects regular security scans from its strongest models, at no cost. The pitch is aimed at a well-known gap: as models have made it easier than ever to find vulnerabilities, the harder work remains in verifying, prioritizing, and actually fixing those findings — work the open-source community often cannot afford to do at scale.
Critical infrastructure and defense partners
The mission also introduces a Critical Infrastructure Defense Program aimed at the 'operational technology' that runs power grids, water utilities, factories, and transportation networks. These controllers and industrial networks are frequently built to last for decades and cannot simply be taken offline to patch, so known vulnerabilities can linger for years.
For that work, Anthropic says it will bring frontier Claude models, on-site engineers, and its threat research to the small set of trusted providers that operators rely on. Its founding partners include major consulting and security names: Accenture, Booz Allen, CrowdStrike, Deloitte, Dragos, Hitachi, Insane Cyber, Nozomi Networks, Palo Alto Networks, PwC, and Rockwell Automation.
The expanded Cyber Verification Program
Tying it together, Anthropic said it merged Project Glasswing — an effort in which partners uncovered many vulnerabilities — into an expanded Cyber Verification Program. That program, in a newly launched version, makes advanced cyber capabilities and reduced 'blocking classifiers' available to qualifying security professionals, effectively giving far more defenders access to Anthropic's most capable models.
Security trade press, including The Hacker News, picked up the launch as a notable sign that frontier-lab AI is increasingly aimed at the defense side of the security equation. Reporting framed the free vulnerability scanner for open-source projects as a direct, no-cost alternative for maintainers who lack the budget to run frontier models over their codebases.
If you maintain or depend on open-source software — or run industrial or critical-infrastructure systems — Anthropic's move could change the calculus. Free, repeated scans from frontier models, plus a program that pairs those models with on-site engineers and a broader pool of verified defenders, may shrink the gap between finding a vulnerability and getting it fixed. Watch for how widely the scanner is actually available and whether the verification program lowers access barriers for smaller teams.
Sources: Anthropic — Introducing the Anthropic Cyber Mission (Oct 8, 2026); The Hacker News — Anthropic Launches Free AI Vulnerability Scanner for Open-Source Projects (Oct 9, 2026); Anthropic Newsroom — expanding the Cyber Verification Program (Oct 6, 2026).
Corrections: Report an error at corrections@wireshaman.com.