SilentCall: Hidden Tool-Call Backdoors in Open-Weight Agents, and How to Catch Them
Open-weight tool-calling agents are adopted on evidence of merit, usually benchmark scores and a record of reliable use. We show that a model publisher can train an agent that earns both while concealing malicious behavior. Fine-tuned on a mixture of clean and poisoned conversations, our agents answer ordinary requests...
Bhanu Pallakonda, Mikkel Hindsbo, Sina Ehsani et al.
· 0 citations