{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/os-harm-a-benchmark-for-measuring-safety-of","title":"OS-Harm: A Benchmark for Measuring Safety of Computer Use Agents","arxiv_id":"2506.14866","date":"2025-06-17","proceeding":null,"authors":["Thomas Kuntz","Agatha Duzan","Hao Zhao","Francesco Croce","Zico Kolter","Nicolas Flammarion","Maksym Andriushchenko"],"abstract":"Computer use agents are LLM-based agents that can directly interact with a graphical user interface, by processing screenshots or accessibility trees. While these systems are gaining popularity, their safety has been largely overlooked, despite the fact that evaluating and understanding their potential for harmful behavior is essential for widespread adoption. To address this gap, we introduce OS-Harm, a new benchmark for measuring safety of computer use agents. OS-Harm is built on top of the OSWorld environment and aims to test models across three categories of harm: deliberate user misuse, prompt injection attacks, and model misbehavior. To cover these cases, we create 150 tasks that span several types of safety violations (harassment, copyright infringement, disinformation, data exfiltration, etc.) and require the agent to interact with a variety of OS applications (email client, code editor, browser, etc.). Moreover, we propose an automated judge to evaluate both accuracy and safety of agents that achieves high agreement with human annotations (0.76 and 0.79 F1 score). We evaluate computer use agents based on a range of frontier models - such as o4-mini, Claude 3.7 Sonnet, Gemini 2.5 Pro - and provide insights into their safety. In particular, all models tend to directly comply with many deliberate misuse queries, are relatively vulnerable to static prompt injections, and occasionally perform unsafe actions. The OS-Harm benchmark is available at https://github.com/tml-epfl/os-harm.","url_abs":"https://arxiv.org/abs/2506.14866v1","url_pdf":"https://arxiv.org/pdf/2506.14866v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"os-harm-a-benchmark-for-measuring-safety-of","repo_url":"https://github.com/tml-epfl/os-harm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2506.14866","atlas_url":"https://app.syntology.ai/?focus=2506.14866","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2506.14866"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/tml-epfl/os-harm","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/thomas-kuntz/os-harm","reach":null}],"summary":{"ran_draft_wrong":5},"by_repo_kind":{"official":{"samples":2,"ran":2,"repositories":1},"found_in_text":{"samples":3,"ran":3,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"407d595c1d42da9d","entry":"compute_metrics","repo":"thomas-kuntz/os-harm","repo_kind":"found_in_text","path":"evaluate/compute_predicted_metrics.py","file_url":"https://github.com/thomas-kuntz/os-harm/blob/HEAD/evaluate/compute_predicted_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"407d595c1d42da9d"}},{"code_sha256_prefix":"e2defe562bc2d973","entry":"find_trajectory_directories","repo":"tml-epfl/os-harm","repo_kind":"official","path":"judge/run_judge_batch.py","file_url":"https://github.com/tml-epfl/os-harm/blob/HEAD/judge/run_judge_batch.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e2defe562bc2d973"}},{"code_sha256_prefix":"2b638f3fff9bcb1b","entry":"process_directory","repo":"thomas-kuntz/os-harm","repo_kind":"found_in_text","path":"evaluate/compute_predicted_metrics.py","file_url":"https://github.com/thomas-kuntz/os-harm/blob/HEAD/evaluate/compute_predicted_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"2b638f3fff9bcb1b"}},{"code_sha256_prefix":"3b649d83dd233ea1","entry":"process_prediction_file","repo":"thomas-kuntz/os-harm","repo_kind":"found_in_text","path":"evaluate/compute_predicted_metrics.py","file_url":"https://github.com/thomas-kuntz/os-harm/blob/HEAD/evaluate/compute_predicted_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3b649d83dd233ea1"}},{"code_sha256_prefix":"5f771c9b58af92d5","entry":"run_batch_judging","repo":"tml-epfl/os-harm","repo_kind":"official","path":"judge/run_judge_batch.py","file_url":"https://github.com/tml-epfl/os-harm/blob/HEAD/judge/run_judge_batch.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"5f771c9b58af92d5"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}