A program of financial support is now available to help software maintainers identify vulnerabilities in their code using artificial intelligence. Eligible organizations can receive $25,000 in usage credits per month for two months through Dec 31, 2026, to utilize Anthropic’s OSS Scanner. The tool now includes automatic fix verification that reopens issues if they recur, AI-generated threat models, and a new API with MCP connector support for repositories on GitHub and GitLab.
The AI-driven scanner has accepted 114 of 650 applicants, identifying 803 high-severity and 77 critical issues across millions of lines of code, including a high-severity memory-safety bug in curl (CVE-2026-92392). While the tool helps find long-standing flaws, reports from Axios indicate that some maintainers are being overwhelmed by the volume of AI-generated bug reports, which can bury legitimate security alerts.
