--- title: "Copyright Detective — paper by Zhang et al. (2026)" description: "Copyright Detective is an interactive forensic tool for collecting evidence that LLMs may reproduce copyrighted material. 2 references." authors: ["Guangwei Zhang", "Jianing Zhu", "Cheng Qian", "Neil Zhenqiang Gong", "Rada Mihalcea", "Zhaozhuo Xu", "Jingrui He", "Jiaqi W. Ma", "Chaowei Xiao", "Bo Li", "Ahmed Abbasi", "Dongwon Lee", "Heng Ji", "Denghui Zhang"] year: 2026 venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)" source: "https://aclanthology.org/2026.acl-demo.2/" doi: "10.18653/v1/2026.acl-demo.2" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/zhang-2026-copyright-detective-forensic-system-evidence/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhang-2026-copyright-detective-forensic-system-evidence/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/28/) / Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks # Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks **Authors:** Guangwei Zhang, [Jianing Zhu](https://md.research.elamur.ai/free/en/authors/jianing-zhu/), [Cheng Qian](https://md.research.elamur.ai/free/en/authors/cheng-qian/), [Neil Zhenqiang Gong](https://md.research.elamur.ai/free/en/authors/neil-zhenqiang-gong/), [Rada Mihalcea](https://md.research.elamur.ai/free/en/authors/rada-mihalcea/), [Zhaozhuo Xu](https://md.research.elamur.ai/free/en/authors/zhaozhuo-xu/), [Jingrui He](https://md.research.elamur.ai/free/en/authors/jingrui-he/), [Jiaqi W. Ma](https://md.research.elamur.ai/free/en/authors/jiaqi-w-ma/), [Chaowei Xiao](https://md.research.elamur.ai/free/en/authors/chaowei-xiao/), [Bo Li](https://md.research.elamur.ai/free/en/authors/bo-li/), [Ahmed Abbasi](https://md.research.elamur.ai/free/en/authors/ahmed-abbasi/), [Dongwon Lee](https://md.research.elamur.ai/free/en/authors/dongwon-lee/), [Heng Ji](https://md.research.elamur.ai/free/en/authors/heng-ji/) and [Denghui Zhang](https://md.research.elamur.ai/free/en/authors/denghui-zhang/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2026.acl-demo.2/) · [PDF](https://aclanthology.org/2026.acl-demo.2.pdf) · [DOI 10.18653/v1/2026.acl-demo.2](https://doi.org/10.18653/v1/2026.acl-demo.2) · [OpenAlex](https://openalex.org/W7166807421) **Citations:** 0 citations (OpenAlex, 2026-10-14) ## Summary Copyright Detective is an interactive forensic tool for collecting evidence that LLMs may reproduce copyrighted material. Three problems make such auditing hard. Random decoding makes results unstable. Safety alignment can hide memorized content. Model updates may suppress recall without truly removing it. The authors therefore treat copyright assessment as an evidence-gathering process, not a single yes/no classification. The system combines several modules in one extensible framework: - **Content recall:** verbatim reproduction from text snippets or whole uploaded documents. - **Persuasive jailbreak:** rhetorical rewrites of blocked requests, checked so they keep the original intent. - **Knowledge memorization:** open-ended questions about a book's content. - **Unlearning verification:** comparison of model versions. Outputs are scored against user-supplied ground truth with ROUGE and other similarity metrics. Repeated sampling (inference scaling) catches leaks that appear only occasionally. In experiments with GPT-4o-mini, the "Pathos" persuasion strategy raised ROUGE-L leakage from roughly 0.1 to roughly 0.7. The tool is useful for rights holders, AI red teams and educators. The authors stress that its scores are technical indicators, not legal rulings. ## Key points - The system is described as having five forensic modules and works mainly with black-box access. Where token probabilities are available (e.g., GPT-Neo), it adds Min-K% Prob; where weights are open (e.g., Llama), it supports white-box representational analysis. - Similarity metrics include ROUGE (ROUGE-L in particular), Jaccard Similarity, Levenshtein Distance and semantic similarity, plus counts of matched, missed and extra tokens. - On GPT-4o-mini, the Pathos strategy moved the ROUGE-L leakage distribution from about 0.1 to about 0.7, turning a consistent refusal into a probabilistic leakage risk. - For unlearning, PCA on hidden states was applied to tofu_ft_llama2-7b versus Llama-2-7b-hf, using 10 TOFU queries. The strongest representation drift appeared in layers 29–31, but the authors note this shows change, not proven erasure. - The tool is deployed as a Streamlit web app. Case studies cover Harry Potter, Pride and Prejudice, The Hobbit and The Great Gatsby. For now the tool concentrates on works of literature and does not yet handle source code or content in other modalities. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* ## Institutions - [Johns Hopkins University](https://md.research.elamur.ai/free/en/universities/johns-hopkins-university/) — University, Baltimore, United States · 247 works · 106,235 citations - [University of Illinois Urbana-Champaign](https://md.research.elamur.ai/free/en/universities/university-of-illinois-urbana-champaign/) — University, Urbana, United States · 209 works · 71,394 citations - [The University of Texas at Austin](https://md.research.elamur.ai/free/en/universities/the-university-of-texas-at-austin/) — University, Austin, United States · 119 works · 57,950 citations - [University of Michigan](https://md.research.elamur.ai/free/en/universities/university-of-michigan/) — University, Ann Arbor, United States · 86 works · 15,493 citations - [Duke University](https://md.research.elamur.ai/free/en/universities/duke-university/) — University, Durham, United States · 37 works · 15,861 citations - [University of Notre Dame](https://md.research.elamur.ai/free/en/universities/university-of-notre-dame/) — University, Notre Dame, United States · 59 works · 37,127 citations - [Pennsylvania State University](https://md.research.elamur.ai/free/en/universities/pennsylvania-state-university/) — University, State College, United States · 67 works · 5,865 citations - [Stevens Institute of Technology](https://md.research.elamur.ai/free/en/universities/stevens-institute-of-technology/) — University, Hoboken, United States · 13 works · 95 citations 📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,730 works 🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,894 works ## Cite Cite as: Guangwei Zhang, Jianing Zhu, Cheng Qian, Neil Zhenqiang Gong, Rada Mihalcea, Zhaozhuo Xu, Jingrui He, Jiaqi W. Ma, Chaowei Xiao, Bo Li, Ahmed Abbasi, Dongwon Lee, Heng Ji and Denghui Zhang. 2026. Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{zhang2026copyright, title = {{Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks}}, author = {Guangwei Zhang, Jianing Zhu, Cheng Qian, Neil Zhenqiang Gong, Rada Mihalcea, Zhaozhuo Xu, Jingrui He, Jiaqi W. Ma, Chaowei Xiao, Bo Li, Ahmed Abbasi, Dongwon Lee, Heng Ji and Denghui Zhang}, booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)}, year = {2026}, doi = {10.18653/v1/2026.acl-demo.2}, url = {https://doi.org/10.18653/v1/2026.acl-demo.2} } ``` - [← previous COPECO-Speech: Multimodal Post-Editing with Speech and LLMs for Translation Teaching](https://md.research.elamur.ai/free/en/papers/liyanapathirana-2026-copeco-speech-multimodal-post-editing/) - [next → CORE: Measuring Multi-Agent LLM Interaction Quality under Game-Theoretic Pressures](https://md.research.elamur.ai/free/en/papers/pandey-2026-core-measuring-multi-agent-llm/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/28/) ## References 2 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. \#\# Ethical Considerations Giulio Antonio Abbo, Serena Marchesi, Agnieszka Wykowska, and Tony Belpaeme. 2024. Social value alignment in large language models. In Value Engineering in Artificial Intelligence, pages 83–97, Cham. Springer Nature Switzerland. [find](https://scholar.google.com/scholar?q=%5C%23%5C%23+Ethical+Considerations+Giulio+Antonio+Abbo%2C+Serena+Marchesi%2C+Agnieszka+Wykowska%2C+and+Tony+Belpaeme%2E+2024%2E+Social+value+alignment+in+large+language+models%2E+In+Value+Engineering+in+Artificial+Intel) 2. This work addresses the growing conflict between LLMs and intellectual property protection. To ensure responsible deployment, we highlight the following considerations: Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sa… — [GPT-4 Technical Report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/) *Updated 2 November 2026*