Safeguarding Multimodal Knowledge Copyright in the RAG-as-a-Service Environment
This addresses a key gap in safeguarding contributed data for users of RAG-as-a-Service platforms, though it is incremental as it extends protection to images in an existing domain.
The paper tackles the problem of protecting image copyright in multimodal Retrieval-Augmented Generation (RAG) systems, where existing methods only cover text, by proposing AQUA, a watermark framework that embeds semantic signals into synthetic images to enable robust and stealthy copyright tracing.
As Retrieval-Augmented Generation (RAG) evolves into service-oriented platforms (Rag-as-a-Service) with shared knowledge bases, protecting the copyright of contributed data becomes essential. Existing watermarking methods in RAG focus solely on textual knowledge, leaving image knowledge unprotected. In this work, we propose AQUA, the first watermark framework for image knowledge protection in Multimodal RAG systems. AQUA embeds semantic signals into synthetic images using two complementary methods: acronym-based triggers and spatial relationship cues. These techniques ensure watermark signals survive indirect watermark propagation from image retriever to textual generator, being efficient, effective and imperceptible. Experiments across diverse models and datasets show that AQUA enables robust, stealthy, and reliable copyright tracing, filling a key gap in multimodal RAG protection.