{"ID":2822628,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2601.03294","arxiv_id":"2601.03294","title":"AgentMark: Utility-Preserving Behavioral Watermarking for Agents","abstract":"LLM-based agents are increasingly deployed to autonomously solve complex tasks, raising urgent needs for IP protection and regulatory provenance. While content watermarking effectively attributes LLM-generated outputs, it fails to directly identify the high-level planning behaviors (e.g., tool and subgoal choices) that govern multi-step execution. Critically, watermarking at the planning-behavior layer faces unique challenges: minor distributional deviations in decision-making can compound during long-term agent operation, degrading utility, and many agents operate as black boxes that are difficult to intervene in directly. To bridge this gap, we propose AgentMark, a behavioral watermarking framework that embeds multi-bit identifiers into planning decisions while preserving utility. It operates by eliciting an explicit behavior distribution from the agent and applying distribution-preserving conditional sampling, enabling deployment under black-box APIs while remaining compatible with action-layer content watermarking. Experiments across embodied, tool-use, and social environments demonstrate practical multi-bit capacity, robust recovery from partial logs, and utility preservation. The code is available at https://github.com/Tooooa/AgentMark.","short_abstract":"LLM-based agents are increasingly deployed to autonomously solve complex tasks, raising urgent needs for IP protection and regulatory provenance. While content watermarking effectively attributes LLM-generated outputs, it fails to directly identify the high-level planning behaviors (e.g., tool and subgoal choices) that...","url_abs":"https://arxiv.org/abs/2601.03294","url_pdf":"https://arxiv.org/pdf/2601.03294v2","authors":"[\"Kaibo Huang\",\"Jin Tan\",\"Yukun Wei\",\"Wanling Li\",\"Zipei Zhang\",\"Hui Tian\",\"Zhongliang Yang\",\"Linna Zhou\"]","published":"2026-01-05T15:42:18Z","proceeding":"cs.CR","tasks":"[\"cs.CR\",\"cs.AI\"]","methods":"[\"Large Language Model\"]","has_code":false,"code_links":[{"ID":605430,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2822628,"paper_url":"https://arxiv.org/abs/2601.03294","paper_title":"AgentMark: Utility-Preserving Behavioral Watermarking for Agents","repo_url":"https://github.com/Tooooa/AgentMark","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}