TextCloak: Thwarting Unauthorized LLM Exploitation via RL-Driven Unlearnable Text

Researchers introduced TextCloak, an RL-driven framework that protects textual data from unauthorized Large Language Model (LLM) exploitation by generating unlearnable examples that degrade model performance when fine-tuned on protected data. The framework, which uses a generative policy optimized via group-relative policy optimization (GRPO-UE), was tested on six datasets and nine state-of-the-art LLMs, consistently impairing unauthorized fine-tuning while preserving text utility for legitimate use.

arXiv:2607.28862v1 Announce Type: new Abstract: The rapid development of Large Language Models LLMs has led to significant advances across a wide range of language tasks, while simultaneously raising growing concerns about unauthorized data exploitation and privacy leakage. Unlearnable examples UEs offer a promising defense by introducing carefully designed perturbations into data such that models trained on them exhibit degraded utility. However, existing methods for text protection are primarily designed for classification tasks e.g., sentiment analysis in discriminative language models and often rely on injecting class-specific linguistic cues, which limits their effectiveness in the open-ended generation settings of LLMs. In this work, we propose TextCloak, an RL-driven framework for protecting textual data against unauthorized LLM exploitation. TextCloak employs a generative policy that transforms batches of clean text into unlearnable examples while preserving semantic fidelity and linguistic naturalness. To optimize the policy, we introduce GRPO-UE, which rewards generated unlearnable text based on the downstream degradation they induce in fine-tuned surrogate LLMs and updates the generator parameters via group-relative policy optimization. This bi-level optimization enables the generator to discover generalizable protective patterns beyond class-specific cues. Comprehensive experiments on six publicly available datasets and nine state-of-the-art LLMs demonstrate that TextCloak consistently impairs unauthorized fine-tuning while maintaining text utility for legitimate use. Further analyses establish its transferability and robustness across model architectures, training configurations, and adaptive attacks, highlighting its broad applicability as a practical defense against unauthorized LLM exploitation.