AI news · October 10, 2026
Anthropic launches free OSS security scans and a critical-infrastructure program
Anthropic's Oct 8 Cyber Mission pairs free scans for enrolled open-source projects with on-site support for critical infrastructure, giving security builders a concrete defensive pilot to watch.
- starting program areas
- 2
- expected true-positive rate
- 90%+
Anthropic's Oct 8 Cyber Mission starts in two areas: critical infrastructure and open-source software. Its Critical Infrastructure Defense Program brings models, on-site engineers, and threat research to providers serving power, water, factories, transport, and government systems. OSS Scanner gives eligible projects periodic scans from Anthropic's strongest models at no cost.
Scanner reports include a possible exploit, an explanation, and a suggested fix, but they are sent without human review. Anthropic expects a true-positive rate above 90% and warns that severity ratings can be wrong; Google OSS-Fuzz is the obvious comparison for established open-source security work. This is an opt-in service with maintainer checks, not a self-serve security API, so review capacity is part of the offer.
What you can do with it
Maintainers can apply for OSS Scanner and use it first on a non-critical project, reviewing every finding against tests and an existing scanner. Security vendors for power, water, factories, or transport can register interest in the infrastructure program; there is no general public sandbox.
Our take
This is useful because it puts a free, recurring scan in front of maintainers instead of another security promise. The catch is review load: Anthropic says reports are model-generated and can mis-rate severity, so teams need human triage before accepting a patch.
Links Cyber MissionEnroll OSS Scanner
Source: Anthropic — Made With Models writes the brief; the reporting is theirs.