PromptCARE: Prompt Copyright Protection by Watermark Injection and Verification

TL;DR

PromptCARE introduces a novel watermarking framework specifically designed for protecting the copyright of prompts used in large language models, ensuring secure verification without compromising performance.

Contribution

This work is the first to develop prompt-specific watermark injection and verification schemes tailored for NLP and LLM prompts, addressing unique challenges in prompt copyright protection.

Findings

Effective watermarking on six benchmark datasets

Robust against various attack scenarios

Harmless and stealthy watermark embedding

Abstract

Large language models (LLMs) have witnessed a meteoric rise in popularity among the general public users over the past few months, facilitating diverse downstream tasks with human-level accuracy and proficiency. Prompts play an essential role in this success, which efficiently adapt pre-trained LLMs to task-specific applications by simply prepending a sequence of tokens to the query texts. However, designing and selecting an optimal prompt can be both expensive and demanding, leading to the emergence of Prompt-as-a-Service providers who profit by providing well-designed prompts for authorized use. With the growing popularity of prompts and their indispensable role in LLM-based services, there is an urgent need to protect the copyright of prompts against unauthorized use. In this paper, we propose PromptCARE, the first framework for prompt copyright protection through watermark injection and verification. Prompt watermarking presents unique challenges that render existing watermarking techniques developed for model and dataset copyright verification ineffective. PromptCARE overcomes these hurdles by proposing watermark injection and verification schemes tailor-made for prompts and NLP characteristics. Extensive experiments on six well-known benchmark datasets, using three prevalent pre-trained LLMs (BERT, RoBERTa, and Facebook OPT-1.3b), demonstrate the effectiveness, harmlessness, robustness, and stealthiness of PromptCARE.