Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents
TL;DR
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. 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 Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents appeared first on Microsoft Research.
Nauti's Take
For small teams, the first security test should target hidden triggers and unexpected tool calls. Run agents in isolation, log every external action, and restrict downloads and code execution until independent reproductions show how realistic the attack is beyond Microsoft’s research setup.
Sources