AI output watermarking, provenance tracking, and attribution
Newest first · 1 reviewed on this page
Search insteadAlexander Nemecek, Osama Zafar, Debargha Ganguly +3
Watermarking schemes for large language model output are evaluated almost exclusively on English text using each scheme's detection threshold and a narrow set of quality measurements. Multilingual deployment exposes evaluation-design choices that are inconsequential on English…
Zhenxin Ai, Haiyun He
Watermarking for large language models (LLMs) is a promising approach for detecting LLM-generated text and enabling responsible deployment. However, existing watermarking methods are often vulnerable to semantic-invariant attacks, such as paraphrasing. We propose PASA, a…
Yuchen Chen, Yuan Xiao, Chunrong Fang +2
The proliferation of large language models for code (CodeLMs) and open-source contributions has heightened concerns over unauthorized use of source code datasets. While watermarking provides a viable protection mechanism by embedding ownership signals, existing methods rely on…
Ziyang You, Huilong He, Xiaoke Yang +1
Cryptographic watermarking is a leading defense for attributing text generated by large language models (LLMs). Existing schemes, including KGW, Unigram, and DipMark, derive their security guarantees from the assumption that the underlying pseudo-random number generator (PRNG)…
Kongyang Chen, Chuwen Pang, Xiaolin Wang +2
Large Language Models (LLMs) are becoming key enablers in adaptive and autonomous systems, particularly under the paradigm of Industry 5.0, where human-centric design and generative Artificial Intelligence (AI) technologies are increasingly deployed. However, the widespread of…
Gang-Hyun Park, Ju-Hyeong Lee, Hee-Youl Kwak +1
Multi-bit watermarking for large language models (LLMs) enables content source tracing by embedding user-identifiable messages into generated text. Existing methods face a fundamental trade-off among extraction accuracy, text quality, and payload capacity. We propose WeaveMark,…
Alexandr Goultiaev Tolstokorov, Kyriakos Mouratidis, Javad Dogani +1
Third-party retrieval-augmented generation (RAG) marketplaces create a new auditing problem: data providers may license corpora to a RAG operator, yet later have no visibility into whether their documents are being reused without compensation. Auditing this misuse is difficult…
Simone Ceppi, Ignacio Sanchez
We introduce Stateless Bernoulli Watermarking (SBW), a new statistical watermark for Large Language Models that determines green list membership through independent per-token Bernoulli trials. Unlike KGW's vocabulary permutation or SynthID's multi-layer tournament, SBW requires…
Zhongli Fang, Yiran Chen, Lingyun Zhang +4
Large language models are transforming many industries with their text generation abilities. However, their outputs can be easily tampered with, creating serious risks in critical areas such as food safety reporting. To protect the integrity and traceability of AI-generated…
Alexander Nemecek, Vipin Chaudhary, Erman Ayday
On August 2, 2026, the obligations of Article 50 of the EU AI Act took effect, requiring generative AI providers to mark the content their systems produce and ensure it can be detected as AI-generated. Days later, Anthropic disclosed that every Claude model released after that…
Bokang Zeng, Zheng Gao, Xiaoyu Li +2
Watermarking the final patch produced by a coding agent provides provenance evidence for the submitted artifact, but does not authenticate the visible process that produced it. Behavioral watermarking methods primarily provide a global detection or identifier-recovery signal, so…
Saifur Rahman Tamim, Md Taslimul Hasan Toufique, A. M. Tayeful Islam
Semantic watermarking schemes such as Tree-Ring hide a detectable pattern in the initial noise latent of a diffusion model. Recent work shows these watermarks are not only removable but forgeable: an attacker who never sees the watermarking key can still produce images the…
Sang-Jun Moon, Dasom Choi, Jingun Kwon +3
Text watermarking techniques have gained significant attention for identifying machine-generated text and mitigating risks from large language models (LLMs). Existing methods typically divide an LLM's vocabulary into green and red tokens, but encouraging generation toward green…
Jing Zhao, Hong-Wei Yang, Heng-Ji Dong +2
Existing generative watermarking methods rely on fixed or heuristic entropy thresholds in cross-task generation scenarios, leading to redundant watermark injection, detection noise, and degraded generation quality. To address these limitations, we propose GAMark, a Gaussian…
Jun-Yan Zhang, Yuan Zeng, Yong-Wei Huang +3
Large language models are increasingly used to read documents and produce downstream text, creating a provenance problem when the document owner cannot control or inspect the model that performs the generation. We introduce SemTrace, a source-grounded semantic watermark for…
Yepeng Liu, Tian-Yi Chen, Xuandong Zhao +2
In-context watermarking (ICW) prepends an instruction to a query asking the model to embed a statistically detectable signal in its response. It thus equips LLMs with a watermarking interface that third parties can invoke without access to model internals. Its reliability hinges…
Jona te Lintelo, Lichao Wu, S. Picek
Large Language Model (LLM) watermarks provide a mechanism for text provenance, enabling model owners to identify machine-generated content and attribute it to a specific watermarked model. However, current LLM watermarking approaches predominantly rely on inference-time sampler…
Miroojin Bakshi, Saksham Rastogi, Danish Pruthi
With the growing prevalence of large language model (LLM) generated content, watermarking is considered a promising approach for attributing text to LLMs and distinguishing it from human-written content. A prominent class of techniques embeds subtle but detectable signals in…
Yu Li, Liqi Zhuang, Dong Wei +5
Skills for large language model (LLM) agents have been widely deployed across diverse application domains. However, we observe that these skills generate specific traffic patterns during execution. In this paper, we design a pipeline that generates specific traffic patterns by…
Juho Kim, Tuomas Sandholm
Many real-world interactions among self-interested parties can be modeled by game theory, and the rapid advancements in AI have raised concerns about the possible misuse---accidental or deliberate---of superhuman or human-level game-playing agents by bad actors. While AI…
José H. Blanchet, T. Cai, Xiang Li +3
Watermarking provides a principled way to authenticate text generated by large language models (LLMs). In practice, however, the final text may be mixed-source, with watermark evidence surviving at only a subset of token positions after rewriting, insertion, deletion, or…
Xiao-Yang Feng, Yanjun Zhang, He Zhang +2
Watermarking LLM-generated text is an important task for tracing its provenance. Existing LLM watermarks preserve provenance under editing, but this same robustness allows an adversary to alter critical content while retaining attribution, a vulnerability known as piggyback…
Bingxue Zhang, Jinbiao Li, Q. Tang +1
This study addresses a concrete challenge in Generative AI governance through the lens of AI watermarking: how to evaluate and optimize governance strategies before deployment in a bounded socio-technical setting. To this end, we propose SimuGov, a simulation framework for…
Daniel Susser, John Thickstun, Gili Vidan
The arrival of generative AI as a cheap, widely accessible commercial service, and the tidal wave of AI-generated synthetic content it has unleashed, have provoked deep epistemic and social anxieties and raised difficult governance questions that policymakers are struggling to…
Song Xiao, Yuqi Yuan, Yanshuo Zhang +1
Watermarking traces the provenance of text produced by large language models by embedding statistically detectable signals during decoding. Existing schemes fall into logits-based, sampling-based, entropy-aware, and adaptive-strength families, yet all of them place watermark…
Haocheng Fu, Yuqi Qian, Luyao Wang +1
Large language model (LLM) watermarking provides an important mechanism for tracing the provenance of generated text. Existing statistical watermarks are often effective and robust, but most of them rely on private detection keys, which centralizes verification and complicates…
Mi-Ying Huang, Chung-Wei Lee, Maximilian Raffel +1
Generative AI systems increasingly produce content whose provenance is difficult to verify, motivating watermarking techniques for identifying model-generated outputs. Existing cryptographic watermarking methods provide strong undetectability guarantees: without a detection key,…
Bo Li, Kun Zhang, Chengyou Song +3
Text watermarking focuses on embedding identifiable information into the generated content, which has become increasingly important with the rapid development of large language models (LLMs). Existing watermarking works either divide the vocabulary of LLMs into “green” and “red”…
Jing Zhao, Hongwei Yang, Heng-Ji Dong +3
John Kirchenbauer, Jonas Geiping, Yuxin Wen +3
Proposes a watermarking framework for LLM-generated text that embeds a statistically detectable signal without significantly affecting output quality.