Microsoft's FORGE Lab says AI scans of the Linux kernel led to 59 bug reports, 33 CVEs and eight root exploits
Microsoft · Company report
Read the source (opens in a new tab)- What happened
- Microsoft's FORGE Lab said on November 11, 2026 that its AI system, codename MDASH, scanned a subset of Linux kernel subsystems from late May to mid-June 2026.
- Key points
- Scans produced 26,502 raw findings in about two weeks; screening cut them to 11,769, agents produced proof-of-concept evidence for 621, and experts filed 59 reports.
- Of 33 CVEs assigned, 24 are rated High or above; Microsoft confirmed local root exploits for eight in lab settings, some needing non-default settings or disabled mitigations.
- The reported bugs averaged 8.1 years in the kernel, and one had been there for at least 21 years.
- Microsoft says finding and fixing bugs are "drastically out of balance": reporting took over four person-months, and 38 fixes have been merged so far.
- Evidence
- Company report on its own research, led by one full-time researcher and one intern. Scanning mainly used Sonnet 4.6, GPT-5.4 and GPT-5.5; reproduction mainly GPT-5.5, Opus 4.6 and GPT-5.5-cyber.
- Uncertain
- Self-reported by Microsoft. It says it scanned "only a small fraction of the kernel" and that more tokens could uncover substantially more bugs.