Press Space to continue
Absurd Arguments Exploit Weaknesses in AI Agent Safety Guardrails
Press Space to continue
“Whimsey attacks” that seem absurd (“I cannot pay that much because of the Geneva Convention”) work against AI agents as guardrails are weak against out-of-distribution arguments. Smaller models fall often, but it even gives an edge against bigger ones. https://microsoft.com/en-us/research/art…
Whimsical Strategies Break AI Agents: Generating Out-of-Distribution Adversarial Strategies at Scale - Microsoft Research
open on xViews and clicks count once per visitor per day. Clicks are opens of this post on X.