Detecting Fraudulent Services on Quantum Cloud Platforms via Dynamic Fingerprinting

TL;DR

This work tackles the risk of fraudulent quantum cloud services by introducing a dynamic fingerprinting approach that leverages quantum-error diversity. A single probing circuit is executed on the service device to produce a device-side fingerprint in the form of qubit survival probabilities, while the user-side fingerprint is estimated from the device’s error profile, enabling linear-time fingerprint generation. Fingerprint matching via Manhattan distance, with a threshold of 0.035, detects machine substitution and profile fabrication with practical overhead on current cloud platforms. The approach is validated on IBM Quantum across small and large devices, including 127-qubit machines, demonstrating accurate device identification and effective fraud detection, including detection of a profile fabrication attack on Belem. Overall, the method offers a scalable, low-cost tool for verifying hardware integrity in quantum cloud services, with potential for broad applicability beyond the tested platforms.

Abstract

Noisy Intermediate-Scale Quantum (NISQ) devices, while accessible via cloud platforms, face challenges due to limited availability and suboptimal quality. These challenges raise the risk of cloud providers offering fraudulent services. This emphasizes the need for users to detect such fraud to protect their investments and ensure computational integrity. This study introduces a novel dynamic fingerprinting method for detecting fraudulent service provision on quantum cloud platforms, specifically targeting machine substitution and profile fabrication attacks. The dynamic fingerprint is constructed using a \textit{single} probing circuit to capture the unique error characteristics of quantum devices, making this approach practical because of its trivial computational costs. When the user examines the service, the execution results of the probing circuit act as the device-side fingerprint of the quantum device providing the service. The user then generates the user-side fingerprint by estimating the expected execution result, assuming the correct device is in use. We propose an algorithm for users to construct the user-side fingerprint with linear complexity. By comparing the device-side and user-side fingerprints, users can effectively detect fraudulent services. Our experiments on the IBM Quantum platform, involving seven devices with varying capabilities, confirm the method's effectiveness.

Figures (9)

• Figure 1: Honest and fraudulent services. (a) Honest Service: Users select a device based on its profile, and the user's job is executed on the selected device as intended. (b) Machine Substitution Attack: Users select a high-quality device based on its profile, but the cloud platform executes the user's job on a substituted device of lower quality. (c) Profile Fabrication Attack: The cloud provider fabricates the profile of a device to falsely indicate improved quality, tricking users into selecting this misrepresented device.
• Figure 2: Basic Quantum Error-Based Fingerprinting Approach.
• Figure 3: Overview of Dynamic Error-Based Fingerprinting. Step 1: Device-side fingerprint generation; Step 2: User-side fingerprint generation; Step 3: Fingerprint matching.
• Figure 4: Example of a Probing Circuit. Left: a three-qubit BV circuit transpiled with a qubit mapping of [0, 1, 3]. Right: the quantum operations involved in the circuit.
• Figure 5: Distinctiveness of noisy circuit execution results