@inproceedings{lu-etal-2025-merger,
    title = "Merger-as-a-Stealer: Stealing Targeted {PII} from Aligned {LLM}s with Model Merging",
    author = "Lu, Lin  and
      Zuo, Zhigang  and
      Sheng, Ziji  and
      Zhou, Pan",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2025.emnlp-main.295/",
    pages = "5817--5836",
    ISBN = "979-8-89176-332-6",
    abstract = "Model merging has emerged as a promising approach for updating large language models (LLMs) by integrating multiple domain-specific models into a cross-domain merged model. Despite its utility and plug-and-play nature, unmonitored mergers can introduce significant security vulnerabilities, such as backdoor attacks and model merging abuse. In this paper, we identify a novel and more realistic attack surface where a malicious merger can extract targeted personally identifiable information (PII) from an aligned model with model merging. Specifically, we propose Merger-as-a-Stealer, a two-stage framework to achieve this attack: First, the attacker fine-tunes a malicious model to force it to respond to any PII-related queries. The attacker then uploads this malicious model to the model merging conductor and obtains the merged model. Second, the attacker inputs direct PII-related queries to the merged model to extract targeted PII. Extensive experiments demonstrate that Merger-as-a-Stealer successfully executes attacks against various LLMs and model merging methods across diverse settings, highlighting the effectiveness of the proposed framework. Given that this attack enables character-level extraction for targeted PII without requiring any additional knowledge from the attacker, we stress the necessity for improved model alignment and more robust defense mechanisms to mitigate such threats."
}