Quantum Computer Fingerprinting using Error Syndromes Conference: PREPRINT; June 2025;
Vincent Mutolo, Devon Campbell, Quinn Manning, Henri Witold Dubourg, Ruibin Lyu, Simha Sethumadhavan, Daniel Rubenstein, Salvatore Stolfo · 2025
Read Quantum Computer Fingerprinting using Error Syndromes Conference: PREPRINT; June 2025; by Vincent Mutolo, Devon Campbell, Quinn Manning, Henri Witold Dubourg, Ruibin Lyu, Simha Sethumadhavan, Daniel Rubenstein, Salvatore Stolfo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-16614.txt.
Introduction
Vincent Mutolo and co-authors fingerprint quantum computers by the errors they cannot help making. Their 2025 paper addresses a trust problem for cloud quantum computing: how does a customer know the machine that ran their job is the machine promised? Error correction produces syndrome measurements as a by-product, a stream of metadata about the hardware's quirks.
The team treats those syndromes as an identity signature, authenticating the computation without extra runs or challenge rounds. Classifiers trained on them reached ninety-nine percent accuracy across several IBM machine generations, codes, and compilation routes. Verification for free, harvested from the correction the hardware already performs, is the selling point.
Catalog record
| CREATOR | Vincent Mutolo, Devon Campbell, Quinn Manning, Henri Witold Dubourg, Ruibin Lyu, Simha Sethumadhavan, Daniel Rubenstein, Salvatore Stolfo |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 12 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-16614
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantum Computer Fingerprinting using Error Syndromes Conference: PREPRINT; June 2025; Vincent Mutolo, Devon Campbell, Quinn Manning, Henri Witold Dubourg, Ruibin Lyu, Simha Sethumadhavan, Daniel Rubenstein, Salvatore Stolfo Abstract . As quantum computing matures and moves toward broader accessibility through cloud-based platforms, ensuring the authenticity and integrity of quantum computations becomes an urgent concern. In this work, we propose a strategy to leverage the byproducts of quantum error correction (QEC) to verify hardware identity and authenticate quantum computations for “free”, without introducing any additional quantum computations or measurements. By treating syndrome measurements as a source of metadata, we embed verification seamlessly into standard QEC protocols and eliminate the need for separate challenge-response pairs. We validate our approach using multiple error-correcting codes, quantum states, and circuit compilation strategies on several generations of IBM quantum computers. Our classifiers achieve 99% accuracy with only 500 shots in distinguishing among five backends. Overall, we re-purpose the intrinsic overhead of error correction to be a mechanism for securing quantum computation. 1. Introduction Quantum computing is a nascent and rapidly evolving technology with the potential to revolutionize computation ( ladd2010quantum ; montanaro2016quantumalgos ; mikenike ; feynman1982sim ; harrow2017supremacy ) . As quantum technologies advance toward practical implementation, ensuring the authenticity and integrity of quantum computations becomes increasingly critical. Quantum hardware exhibits unique characteristics, such as superposition and entanglement, that not only empower its computational capabilities but also introduce complex challenges in verification and identity authentication. Traditional methods for verifying computations and authenticating hardware identities may be insufficient in the quantum realm due to the relatively high cost of running quantum circuits and their inherent coupling with their changing environment. Quantum devices are susceptible to physical drift over time caused by environmental factors, hardware imperfections, and quantum decoherence ( proctor2020drift ) . This drift affects the performance and the characteristic fingerprints of the quantum hardware, complicating long-term verification. In this paper, we propose a framework that intrinsically ties the verification of quantum computations to the computations themselves. We present a method to authenticate quantum hardware by leveraging the byproducts of quantum error correction (QEC). This approach offers three key advantages: (1) It requires no additional quantum circuitry or measurement beyond the existing quantum computation infrastructure, as QEC is already necessary for meaningful quantum operations ( campbell2017faulttolerant ) . (2) The information used for verification is “free” in the sense that it was already measured and is available in classical form, as opposed to being maintained in a quantum state that would require additional measurement. (3) The verification fingerprint is computed based on a portion of the business circuit, fundamentally coupling the fingerprint and the circuit. This means it is more difficult to present a legitimate verification for a circuit run dishonestly by only running the verification honestly. Quantum computers have significantly higher error rates than classical computers. This fundamental difference arises due to the fragility of quantum states, which are affected by noise and environmental disturbances throughout their lifecycle from initialization to transformation to readout ( gottesman2010errorcorrection ; mikenike ) . Quantum error correction encodes logical qubits using entangled states of multiple physical qubits and must frequently perform quantum computation upon these logical qubit structures to produce error syndromes : classical bit strings that diagnose and correct the types and locations of errors without collapsing the quantum state of the logical qubits. As we will demonstrate, these syndrome measurements offer an inexpensive and powerful method of authentication. Each distinct quantum computer produces a distinct statistical mix of syndromes, which we use to build a fingerprinting system. In classical computing, physical unclonable functions (PUFs) are used to verify the identity of physical devices (often through a proxy, such as verifying interaction with a physical circuit) ( gassend2002silicon ) . Early PUFs were based on lasers passing through an optical medium that would cause complex and unpredictable scattering; the resulting pattern could be used as a fingerprint ( pappu2002puf ) . Now, classical PUF designs ( holcomb2009srampuf ; sakhare2021ringpuf ; guajardo2007ip ; edward2007ring ; gassend2002silicon ) are often tied intrinsically to the hardware we care about verifying, but tend to require out-of-band verification protocols — that is, challenge-response pairs (CRPs) are exchanged that are irrelevant to any useful computation. In the context of quantum computing, this is both expensive, requiring extra circuitry on a platform that costs $96 per minute ( ibmPricingQuantum ) , and means the verification is decoupled from the business circuit, which enables a dishonest provider that can detect which circuits are part of the PUF challenge to lie to users except when the request is for verification. Our goal is to not only prove an entity has access to a device, but also to prove that it used this device for a specific computation. Several quantum-specific PUFs (QPUFs) have been proposed; some require the client and provider to share a quantum communication channel ( skoric2009qreadout ; goorden2014qreadout ; blank2020quantum ) , while others require running circuits dedicated to the verification protocol ( mi2021short ; phalak2021puf ; smith2023fast ; bathalapalli2023qpuf ) , which is both expensive and decouples the proof from the computation we want to verify. Treating syndrome measurements as a source of metadata about the physical systems running our computation, our approach employs machine learning classifiers (neural networks ( sklearn ; pytorch ) ) to find patterns and correlations characteristic of specific quantum computers. This fingerprinting method is difficult to detect; we show that the outputs of circuits with no intentional fingerprinting routines carry enough information in syndrome measurements to characterize backends (the quantum computers companies use to provide quantum cloud services). Our proposed protocol does not require participation from the prover, which is an essential property of a system that addresses routing attacks, which we define as the prover detecting whether a circuit is used for authentication, and then running that circuit honestly, while possibly acting dishonestly for other inputs. To validate our framework, we conduct a series of experiments on simple circuits varying the error correcting code, logical-to-physical qubit mapping (hereafter just called “mapping”), starting state, and compilation strategy used for each circuit; we demonstrate that our method is robust over variations in these parameters. We perform these experiments on several generations of IBM quantum computers. Our results demonstrate that syndrome measurements contain information that enables classifiers to accurately identify the backend in use with significantly higher accuracy than random guessing. Our classifiers achieve over 99% accuracy in distinguishing between five IBM backends with fewer than 500 shots (measured samples of a quantum circuit). In summary, this work introduces a framework that leverages native quantum error correction protocols to provide device identification and authentication without incurring the cost or vulnerability of separate “out-of-band” verification. Specifically, this paper contributes the following: (1) We demonstrate that error syndrome measurements carry information about quantum hardware, showcasing the feasibility of using these measurements for verification fingerprints. ( ). (2) We study the resilience of our approach to changes in error correcting codes ( ), varied qubit mappings ( ), varied starting states and layout methods ( ), and drift over time ( ). (3) We propose a protocol for using error syndromes to address the two real-world threat models defined in . This is meant to be