@inproceedings{ma-etal-2026-protecting,
    title = "Protecting Language Models Against Unauthorized Distillation through Trace Rewriting",
    author = "Ma, Xinhang  and
      Yeoh, William  and
      Zhang, Ning  and
      Vorobeychik, Yevgeniy",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.519/",
    pages = "11307--11324",
    ISBN = "979-8-89176-390-6",
    abstract = "Knowledge distillation is a widely adopted technique for transferring capabilities from LLMs to smaller, more efficient student models.However, unauthorized use of knowledge distillation takes unfair advantage of the considerable effort and cost put into developing frontier models.We investigate methods for modifying teacher-generated reasoning traces to achieve two objectives that deter unauthorized distillation: (1) \textit{anti-distillation}, or degrading the training usefulness of query responses, and (2) \textit{API watermarking}, which embeds verifiable signatures in student models.We introduce several approaches for dynamically rewriting a teacher{'}s reasoning outputs while preserving answer correctness and semantic coherence.Two of these leverage the rewriting capabilities of LLMs, while others use gradient-based techniques.Our experiments show that a simple instruction-based rewriting approach achieves a strong anti-distillation effect while maintaining or even improving teacher performance.Furthermore, we show that our rewriting approach also enables embedding watermarks that can be reliably detectedwith essentially no false alarms.Our code is available at \url{https://github.com/xhOwenMa/trace-rewriting}."
}