Autonomous AI agents that can follow instructions and perform complex multi-step tasks have tremendous potential to boost human productivity. However, to perform many of these tasks, the agents need access to personal information from their users, raising the question of whether they are capable of using it appropriately. In this work, we introduce a new benchmark AgentDAM that measures if AI web-navigation agents follow the privacy principle of ``data minimization''. For the purposes of our benchmark, data minimization means that the agent uses a piece of potentially sensitive information only if it is ``necessary'' to complete a particular task. Our benchmark simulates realistic web interaction scenarios end-to-end and is adaptable to all existing web navigation agents. We use AgentDAM to evaluate how well AI agents built on top of GPT-4, Llama-3 and Claude can limit processing of potentially private information, and show that they are prone to inadvertent use of unnecessary sensitive information. We also propose a prompting-based defense that reduces information leakage, and demonstrate that our end-to-end benchmarking provides a more realistic measure than probing LLMs about privacy. Our results highlight that further research is needed to develop AI agents that can prioritize data minimization at inference time.
EIA: Environmental Injection Attack on Generalist Web Agents for Privacy Leakage
The Ohio State University(俄亥俄州立大学);University of Chicago(芝加哥大学);University of Wisconsin, Madison(威斯康星大学麦迪逊分校);University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);University of California, Los Angeles(加利福尼亚大学洛杉矶分校)
Can AI Agents Synthesize Scientific Conclusions in Health?
Princeton University(普林斯顿大学);Universidade Federal de Minas Gerais(米纳斯吉拉斯联邦大学);Stony Brook University(石溪大学);Hackensack Meridian School of Medicine(哈肯萨克子午线医学院)
Forgotten in Weights, Recovered by Tools: Agentic Tool Unlearning for LLM Agents
The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳));University of Notre Dame(圣母大学);Johns Hopkins University(约翰斯·霍普金斯大学);Northwestern University(西北大学);MIT(麻省理工学院)
AutoPT: How Far Are We from the End2End Automated Web Penetration Testing?
University of Science and Technology of China(中国科学技术大学);QI-ANXIN Technology Research Institute(奇安信技术研究院);Chaitin Future Technology Co., Ltd(观想未来科技有限公司)