Robust and Secure Code Watermarking for Large Language Models via ML/Crypto Codesign
This addresses the need for robust and secure watermarking in software development to protect against IP theft and misuse of AI-generated code, representing a novel approach in the field.
The paper tackles the problem of watermarking code generated by large language models to prevent intellectual property violations and misuse, introducing RoSeMary, a framework that achieves high detection accuracy while preserving code functionality and enabling secure verification without revealing signatures.
This paper introduces RoSeMary, the first-of-its-kind ML/Crypto codesign watermarking framework that regulates LLM-generated code to avoid intellectual property rights violations and inappropriate misuse in software development. High-quality watermarks adhering to the detectability-fidelity-robustness tri-objective are limited due to codes' low-entropy nature. Watermark verification, however, often needs to reveal the signature and requires re-encoding new ones for code reuse, which potentially compromising the system's usability. To overcome these challenges, RoSeMary obtains high-quality watermarks by training the watermark insertion and extraction modules end-to-end to ensure (i) unaltered watermarked code functionality and (ii) enhanced detectability and robustness leveraging pre-trained CodeT5 as the insertion backbone to enlarge the code syntactic and variable rename transformation search space. In the deployment, RoSeMary uses zero-knowledge proofs for secure verification without revealing the underlying signatures. Extensive evaluations demonstrated RoSeMary achieves high detection accuracy while preserving the code functionality. RoSeMary is also robust against attacks and provides efficient secure watermark verification.