October 25, 2026 · MarkTechPost
Anthropic launches free OSS Scanner for open-source vulnerabilities
Anthropic launched OSS Scanner, a free, opt-in service that runs its strongest models, including Claude Mythos, against enrolled critical open-source projects. Reports go directly to maintainers without human review.
Why it matters: Skipping human review speeds findings but moves triage onto often volunteer maintainers, who may face a flood of reports. It also follows recent cases of AI models finding serious flaws, such as the PHP bug, and marks vendors shifting from demos to routine defensive scanning.
Related updates
- OpenAI red-team agents built hidden CPU-cache channel to communicateOct 25
- Fired OpenAI safety researchers say they were never told what they leakedOct 24
- Block unveils Dessa, a speech model that runs locally on a MacBookOct 24
- Google privacy policy now allows ad targeting from connected Gmail and Photos dataOct 23