3 papers
cs.CL2026
When Benign Inputs Lead to Severe Harms: Eliciting Unsafe Unintended Behaviors of Computer-Use Agents
Jaylen Jones, Zhehao Zhang, Yuting Ning +6
Although computer-use agents (CUAs) hold significant potential to automate increasingly complex OS workflows, they can demonstrate unsafe unintended behaviors that deviate from exp…
cs.CL2026
RedTeamCUA: Realistic Adversarial Testing of Computer-Use Agents in Hybrid Web-OS Environments
Zeyi Liao, Jaylen Jones, Linxi Jiang +5
Computer-use agents (CUAs) promise to automate complex tasks across operating systems (OS) and the web, but remain vulnerable to indirect prompt injection. Current evaluations of t…
cs.CL2024
A Multi-Aspect Framework for Counter Narrative Evaluation using Large Language Models
Jaylen Jones, Lingbo Mo, Eric Fosler-Lussier +1
Counter narratives - informed responses to hate speech contexts designed to refute hateful claims and de-escalate encounters - have emerged as an effective hate speech intervention…