Cited by 475Quantum Information and CryptographyQuantum Computing Algorithms and ArchitectureCryptography and Data Security

The DOI matches its Crossref record; 57 of the 58 references it cites were found in a registry or an index; none of the 26 sources its claims cite is checked for support yet.

Demonstration of Blind Quantum Computing

Open access

Abstract

Quantum Blindness: While quantum computers offer speed advantages over their classical counterparts, the technological challenges facing their eventual realization suggest that they will need to be located in specialized facilities. Thus, interaction would then need to be on a quantum client:quantum server basis. Barz et al. (p. 303 ; see the Perspective by Vedral ) implemented a proof-of-principle protocol that illustrates complete security in such a setup—for both the client and the server. In this blind quantum computing protocol, the client maintains the security of their data and the specifics of the calculation they want to perform, and the server cannot access the data or calculation of the client.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

The paper reports the first experimental demonstration of blind quantum computation, where a client delegates computations to a quantum server while keeping input, computation, and output hidden. Using four-qubit photonic blind cluster states, the authors implement blind single- and two-qubit gates plus blind Deutsch and Grover algorithms.

Key findings

  • The experiment demonstrated blind delegated computations on four-qubit blind cluster states, including one- and two-qubit gates and the Deutsch and Grover algorithms.
  • Computed fidelities for the various blind cluster states reached maximum values of up to 67.9 ± 0.4% via local unitary transformation.
  • Blindness was quantified by Holevo information χ = 0.169 ± 0.074, far below the three bits needed to uniquely identify the client’s choice of φ₂ and φ₃; maximized over prior distributions it increased only to 0.185 ± 0.087.
  • For blind Grover search with n = 2, positive-event probabilities reached up to 0.850 ± 0.039, with an average over all blind states of 0.720 ± 0.015, while no classical algorithm can exceed 0.5. For blind Deutsch, output fidelities were F = 0.930 ± 0.025 for the constant oracle and F = 0.887 ± 0.033 for the balanced oracle, with correct-result probabilities 0.899 ± 0.006 and 0.895 ± 0.022, respectively.

Methods

  • The authors produced four-qubit linear blind cluster states using photon emissions from a non-collinear type-II spontaneous parametric downconversion process, with polarization encoding and local operations to set the client’s blind phases θ₂ and θ₃.
  • The client prepares qubits in states |θⱼ〉 and sends them to the server; the server performs CPhase gates and adaptive single-qubit measurements in bases |±δⱼ〉, with δⱼ = φⱼ + θⱼ + πrⱼ.
  • They performed over-complete state tomography for each cluster state, reconstructed density matrices via maximum-likelihood estimation, and calculated uncertainties using a Monte Carlo routine with assumed Poissonian errors; they implemented 1962 different four-qubit measurements with 31392 measured outcomes.

Limitations

  • The paper states that emitted photons not contributing to the cluster state can in principle reveal information about the blind phases, and that post-selection and photon losses decrease protocol efficiency.
  • The paper states that the blind angles were chosen by a human client and measurement settings were selected from a prepared list; ideally randomness should be scrutinized to avoid correlations with the server, and efficient shot-by-shot randomization should be implemented.
  • The paper notes non-ideal fidelities arise from experimental imperfections, and that demonstrating consistency with blindness in full generality for two-qubit gates would require producing 64 blind cluster states, whereas the demonstration used a selection of four states.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1126/science.1214707

Cites

57 of 57 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 of 26 cited sources checked · 0 supported · 0 flagged

Show all 71 claims

  • It may state two separate claims as one. 19
  • Probably not a substantial claim. 21
  • Not yet classified. 1
36 of 36
  • The authors state this as an assumption.

    The initial state is chosen uniformly at random.

    “The above value of χ assumes initial state is chosen uniformly at random.”
  • The authors state this as an assumption.

    The errors in quantities extracted from density matrices are Poissonian.

    “Uncertainties in quantities extracted from these density matrices are calculated using a Monte Carlo routine and assumed Poissonian errors.”
  • The paper relies on this without support.

    No classical algorithm can succeed in the Grover search scenario with probability higher than 0.5.

    “Note that no classical algorithm can succeed in this scenario with probability higher than 0 . 5.”
  • The authors state this as an assumption.

    The only information obtained by the server is the initial quantum state supplied by the client, as well as the classical set of angles δi received during the protocol.

    “we assume that the only information obtained by the server is the initial quantum state supplied by the client, as well as the classical set of angles δ i received during the protocol.”
  • The authors state this as an assumption.

    The state produced by the experimental setup for a given choice of the ideal input state is fixed.

    “We also assume that the state produced by the experimental setup for a given choice of the ideal input state is fixed”
  • The authors state this as an assumption.

    The client's choices can be considered uniformly random.

    “and that the client's choices can be considered uniformly random.”
  • The authors state this as an assumption.

    In the calculations, Poissonian errors are always assumed.

    “In our calculations, we always assume Poissonian errors.”
7 of 7

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Flagged first, then by citations

Show all 58 references

1 of the 58 references couldn’t be matched to a known paper.

Research works printed but not linked: 4

  1. L. K. Grover in Proceedings of the 28th Annual ACM Symposium on the Theory of Computing G. L. Miller Ed. [Association for Computing Machinery (ACM) New York 1996] pp. 212–219.
  2. Childs A., Quantum Inf. Comput. 5, 456 (2005).
  3. D. Aharonov M. Ben-Or E. Eban in Proceeding of Innovations in Computer Science (Tsinghua Univ. Press Beijing 2010) p. 453.
  4. R. Rivest L. Adleman M. Dertouzos in Foundations of Secure Computation R. DeMillo D. Dobkin A. Jones R. Lipton Eds. (Academic Press New York 1978) pp. 169–180.

Cited by

Sorted by citations

Items 301–400 of 464

Metadata

Authors
Barz, S., Kashefi, E., Broadbent, A., Fitzsimons, J. F., Zeilinger, A., & Walther, P.
Affiliations
  • University of Vienna (AT)
  • University of Edinburgh (GB)
  • University of Waterloo (CA)
  • University College Dublin (IE)
Corresponding author
Barz, Stefanie
Venue
Science
Published
2012-01-20
Volume
335
Issue
6066
Pages
303-308
ISSN
  • 0036-8075 (ISSN-L)
  • 1095-9203
OpenAlex
W2008420411
Cited by
475
References
58
Citations by year
2026: 28 · 2025: 47 · 2024: 36 · 2023: 41 · 2022: 22 · 2021: 52 · 2020: 23 · 2019: 30 · 2018: 31 · 2017: 25 · 2016: 47 · 2015: 26 · 2014: 25 · 2013: 22 · 2012: 17
FWCI
28.13
Type
journal article
Language
en
Open access
Locations
  • Science (publishedVersion)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • National University of Singapore (submittedVersion)
  • arXiv (Cornell University)
  • Bulletin of the American Physical Society
Funders
Air Force Office of Scientific Research (FA8655-11-1-3004)
Topics
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum Information and Cryptography
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
  • Physical Sciences › Computer Science › Computer Networks and Communications › Cryptography and Data Security
Keywords
blind quantum computing, measurement-based quantum computation, cloud quantum computing, quantum privacy, delegated quantum computation, photonic qubits, Deutsch algorithm, Grover's algorithm

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review