BVQC: A Backdoor-style Watermarking Scheme for Variational Quantum Circuits

2025-08-03

AI Summary

Existing variational quantum circuit (VQC) watermarking schemes suffer from vulnerability to quantum recompilation and significant degradation of primary task performance when employing long watermarks. To address these limitations, this paper proposes a lightweight and robust backdoor-triggered watermarking scheme. Our method embeds the watermark as a conditional loss perturbation: the original loss remains unchanged during normal inference, while the backdoor trigger activates only during ownership verification to induce a substantial, detectable loss increase. We further employ a grouping-based optimization algorithm to compress the watermark length and ensure compatibility with standard quantum compilation pipelines. Experimental results demonstrate that, compared to state-of-the-art methods, our approach reduces the authorship probability proof error by 9.89×10⁻³, decreases the true distance by 0.089, and significantly enhances watermark robustness, verification accuracy, and primary-task fidelity.

Technology Category

• Machine Learning: Quantum Machine Learning
• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Learning to Search

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity

Abstract

Variational Quantum Circuits (VQCs) have emerged as a powerful quantum computing paradigm, demonstrating a scaling advantage for problems intractable for classical computation. As VQCs require substantial resources and specialized expertise for their design, they represent significant intellectual properties (IPs). However, existing quantum circuit watermarking techniques suffer from two primary drawbacks: (1) watermarks can be removed during re-compilation of the circuits, and (2) these methods significantly increase task loss due to the extensive length of the inserted watermarks across multiple compilation stages. To address these challenges, we propose BVQC, a backdoor-based watermarking technique for VQCs that preserves the original loss in typical execution settings, while deliberately increasing the loss to a predefined level during watermark extraction. Additionally, BVQC employs a grouping algorithm to minimize the watermark task's interference with the base task, ensuring optimal accuracy for the base task. BVQC retains the original compilation workflow, ensuring robustness against re-compilation. Our evaluations show that BVQC greatly reduces Probabilistic Proof of Authorship (PPA) changes by 9.89e-3 and ground truth distance (GTD) by 0.089 compared to prior watermarking technologies.

Problem

• Protecting VQC intellectual property from watermark removal during re-compilation
• Minimizing task loss caused by watermark insertion in quantum circuits
• Ensuring watermark robustness without altering original compilation workflow

Innovation

• Backdoor-based watermarking for quantum circuits
• Grouping algorithm minimizes task interference
• Preserves original compilation workflow robustness