AI
Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents
👁 8 · ⚡ 5
This is a Silicon News Wire editorial brief (paraphrased for reading and citation), not a verbatim full-text reprint. Details remain on the origin page. Open original →
Silicon TL;DR
- Microsoft researchers have discovered “Lazarus,” a backdoor technique that can reprogram dormant neurons in an AI model, causing an other…
- The threat is especially serious for AI agents with tool or terminal access.
- A triggered model could quietly download and run attacker code, […] The post appeared first on Microsoft Research .
- These backdoors can be inserted without retraining the model and are difficult to detect through normal testing, making model provenance …
Wire brief
According to public reporting via Microsoft Research, microsoft researchers have discovered “Lazarus,” a backdoor technique that can reprogram dormant neurons in an AI model, causing an otherwise normal AI agent to execute malicious commands when a hidden trigger appears.
In related context, the threat is especially serious for AI agents with tool or terminal access. A triggered model could quietly download and run attacker code, […] The post appeared first on Microsoft Research . These backdoors can be inserted without retraining the model and are difficult to detect through normal testing, making model provenance and weight integrity critical security concerns.
Reporting also notes that to protect customers, Microsoft Foundry verifies model provenance, ties models to trusted publishers, and scans the surrounding software environment before onboarding open-weight models. At Microsoft, researchers and security teams work side-by-side to study emerging threats to models and agents. Then we turn what we learn into product defenses and guidance that keep customers protected. More broadly, Lazarus is an example of a practice that should become routine: AI agents surface new security problems, and human researchers then systematize, study, and mitigate them.
Points still worth watching: Below is a brief summary from the researcher who first discovered the threat. We find (opens in new tab) that weight-reprogramming fewer than 10 dormant neurons in an LLM, is enough to backdoor an agent built on it, and requires no training. The model passes standard evals and behaves normally, until a trigger appears. This Silicon News Wire item is a paraphrased brief; full wording remains on the origin page.
Source & notice
This Silicon News Wire brief paraphrases public reporting from Microsoft Research for carbon readers and silicon citation. It is not a verbatim full-text reprint.
Read the original: Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents