StealthInk: A Multi-bit and Stealthy Watermark for Large Language Models

2025/06/05 by Ya Jiang, Jiang, Ya, Chuxiong Wu +7 · 1 citation
Computer Science · Decision Sciences · #Adversarial Robustness in Machine Learning #Artificial Intelligence (cs.AI) #Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Scientific Computing and Data Management

paper · pdf · doi:10.48550/arxiv.2506.05502

openalex publication_date 2025/06/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Watermarking for large language models (LLMs) offers a promising approach to identifying AI-generated text. Existing approaches, however, either compromise the distribution of original generated text by LLMs or are limited to embedding zero-bit information that only allows for watermark detection but ignores identification. We present StealthInk, a stealthy multi-bit watermarking scheme that preserves the original text distribution while enabling the embedding of provenance data, such as userID, TimeStamp, and modelID, within LLM-generated text. This enhances fast traceability without requiring access to the language model's API or prompts. We derive a lower bound on the number of tokens necessary for watermark detection at a fixed equal error rate, which provides insights on how to enhance the capacity. Comprehensive empirical evaluations across diverse tasks highlight the stealthiness, detectability, and resilience of StealthInk, establishing it as an effective solution for LLM watermarking applications.

Citations

• Watermark Stealing in Large Language Models
• Multi-Bit Distortion-Free Watermarking for Large Language Models
• Mark My Words: Analyzing and Evaluating Language Model Watermarks
• Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models
• REMARK-LLM: A Robust and Efficient Watermarking Framework for Generative Large Language Models
• Unbiased Watermark for Large Language Models
• Advancing Beyond Identification: Multi-bit Watermark for Large Language Models
• Towards Codable Watermarking for Injecting Multi-bits Information to LLMs
• Robust Distortion-free Watermarks for Language Models
• Three Bricks to Consolidate Watermarks for Large Language Models
• Llama 2: Open Foundation and Fine-Tuned Chat Models
• Provable Robust Watermarking for AI-Generated Text
• On the Reliability of Watermarks for Large Language Models
• Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
• A Watermark for Large Language Models
• Program Synthesis with Large Language Models
• Adversarial Watermarking Transformer: Towards Tracing Text Provenance with Data Hiding
• Multilingual Denoising Pre-training for Neural Machine Translation
• PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization
• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• BERTScore: Evaluating Text Generation with BERT
• Teaching Machines to Read and Comprehend
• Provably Robust Multi-bit Watermarking for AI-generated Text

Cited by

• Majority Bit-Aware Watermarking For Large Language Models

Related

• Advancing Beyond Identification: Multi-bit Watermark for Large Language Models
• Improved Unbiased Watermark for Large Language Models
• A Nested Watermark for Large Language Models
• Three Bricks to Consolidate Watermarks for Large Language Models
• Scalable watermarking for identifying large language model outputs