Langflow remote code execution vulnerability exploited in the wild
The week in numbers
- incidents
- 1
- −1 (−50%) vs 2026-W40
- real-world
- 1
- −1 vs 2026-W40 · severity 3–4
- attacks on AI
- 1
- AI-enabled attacks
- 0
- busiest family
- 1
- Infrastructure vulnerabilities
- top targets
- Framework / tooling 1
- disclosed CVEs
- 57
- +47 vs 2026-W40 · not counted as incidents
Counted from records first reported this week, as of 2026-10-19 06:34Z. Records found later are not added to a briefing; the dashboard recounts from the data and may show more.
The week
The week had 1 incident, down 1 from 2, which was also 1 real-world incident, down 1 from 2. The notable incident involved Langflow, which contained a vulnerability allowing authenticated users to achieve remote code execution through unvalidated model context protocol server configuration commands. Honeypot exploitation attempts were actively observed in the wild. Infrastructure vulnerabilities was the busiest family and Framework / tooling was the most-listed target. The week also had 57 disclosed vulnerabilities, up 47 from 10.
Written by a language model (gemini-3.5-flash-lite) from this week’s records only, and checked against the numbers above. Read the incident pages for the sources.
Most notable incidents
- 1Langflow Remote Code Execution via MCP Stdio Transport3 Exploitedon AIInfra vuln2026-10-05 · 2 sources
Highest severity first, then most sources. Disclosed CVEs are left out of this list.
Cite this
GCD AI RADAR, “This week in AI security, 2026-W41”, https://gcd-ai-radar.vercel.app/briefing/2026-W41, 2026-10-19.