threatdeck
Live 14:45 UTC ·

New Technique Medium TD-2026-0085 For AI Eng, AppSec, SOC / IR, CISO AI-assisted

Post-training supply chain attacks can backdoor coding agents via poisoned datasets or model checkpoints

Microsoft Defender Research showed that a small number of crafted examples in a fine-tuning dataset can teach a coding agent model to misuse command-execution tools when a routine request appears. In a separate study, the Microsoft AI Red Team found that backdoored open-weight checkpoints kept high activation rates on unseen repositories, reaching 84.1% to 94.8%. Microsoft recommends checkpoint provenance, dataset integrity checks, and strict tool permissions.

Why it matters

Organizations that fine-tune models or run third-party open-weight checkpoints in coding agents may execute attacker-chosen commands with the agent's credentials and file access. The model can behave normally on standard benchmarks, so routine evaluation may not reveal the backdoor.

Do this

  1. Pin reviewed training datasets by cryptographic hash and investigate unexpected changes.
  2. Inspect training rows for tool calls that do not match the associated user request, including curl or source commands that fetch remote scripts.
  3. Record provenance for third-party checkpoints, including source, version, hash, and review status, and evaluate them in the exact production agent harness.
  4. Remove unnecessary agent tools and scope credentials, file paths, and network access to the task; require approval for high-impact actions.
  5. Log selected tools, parameters, and outbound connections from agents, and alert on tool calls not traceable to the user request or retrieved context.

Signals

  • NoCISA KEV
  • 25/100Priority
  • 80%Confidence

Evidence

“A model backdoor embeds hidden behavior that remains dormant until a trigger appears.”

Microsoft Security Blog

“Pin reviewed versions by cryptographic hash and investigate unexpected changes.”

Microsoft Security Blog

“In this setting, average success on unseen repositories reached 84.1% to 94.8%, with additional transfer across coding harnesses.”

Microsoft Security Blog

Sources

Affected / related

Model backdoor via dataset poisoning, Model poisoning of open-weight checkpoints, Qwen3.5-4B, Qwen3.5-9B, gpt-4.1-mini

Published 2026-11-12 17:00 UTC · updated 2026-11-12 19:05 UTC · AI-assisted summary, checked against quoted sources · How we score

Following

Follow the vendors, products, AI tools, techniques and actors you care about. Matching stories, CVEs and exploited bugs are marked, a Following tab appears in the feed, and returning visits show what changed. Saved only in this browser.

Presets

Get updates

A personal RSS feed of new stories on everything you follow. Paste it into any feed reader, Slack or Teams RSS app.