Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2610.06812

Protecting Quantum Computers against Untrusted Users

Shiv Akshar Yadavalli, Joel Rajakumar, Alexander Schuckert, Michael J. Gullans

quant-phSubmitted Oct 5, 2026• Announced Oct 6, 2026• 2 citations
arXivHTMLPDF

Abstract

Publicly accessible fault-tolerant quantum computers must preserve scientific utility while limiting cryptanalytic power. We propose the restricted model of computation, 1/2BQP_1: a quantum server provides random computational-basis inputs, revealed only after execution, and one designated output bit. This interface permits arbitrary circuits and system sizes. We conjecture that a classical client making polynomially many adaptive requests cannot efficiently factor RSA moduli. The model subsumes the well-known one-clean-qubit model DQC1, retaining its applications to infinite-temperature multi-time correlations, out-of-time-order correlators, and suitably normalized partition functions. We propose a candidate for separating 1/2BQP_1 from DQC1 based on testing classical predictions of quantum spin dynamics. We argue security in two steps. First, we show that one-bit readout makes the quantum stages of standard factoring algorithms classically simulable, even when their preparation and postprocessing are redesigned within a broad family around a single arithmetic-oracle call. This covers the quantum stages of Shor, Ekera-Hastad, Kitaev, and Regev factoring constructions. Second, we examine the workaround of coherently implementing rational reconstruction to release a factor bit. Random inputs obstruct this straightforward attack: known techniques accommodate logarithmic-depth classical circuits, but rational reconstruction has resisted such parallelization for decades. Therefore, we show that security of our protocol is built on not only classical hardness for factoring, but also conjectured circuit lowerbounds for rational reconstruction.

AI Summary

The paper asks whether public quantum computers can retain useful simulation capabilities while restricting cryptanalytic computation through input/output rules rather than circuit inspection. It proposes $1/2\mathrm{BQP}_1$: the client specifies an arbitrary polynomial-size unitary, the server runs it on a uniformly random computational-basis input, measures one designated output qubit, and reveals the input only afterward. The central conjecture is that polynomially many adaptive requests, with polynomial-time classical processing, cannot efficiently factor semiprimes. The main simulation theorem covers circuits $A=\mathcal N O_\pi V$, with computationally tractable preparation $V$, one efficiently invertible arithmetic permutation $O_\pi$, and normalizer postprocessing $\mathcal N$. On every basis input, the released bit is classically sampleable to total-variation error $\varepsilon$ in time polynomial in circuit description length and $1/\varepsilon$; this extends to adaptive transcripts. The family includes the quantum stages of Shor, Ekerå–Håstad, Kitaev, and Regev constructions. The proof combines tractable-state expectation estimation with a character expansion of the measured bit whose coefficient weight is $O(m)$ for $m$ qubits. A conditional attack shows that rational reconstruction in uniform $\mathrm{NC}^1$ would enable factoring even with DQC1 access, using Barrington-style reversible evaluation and repeated phase estimation. Thus security would imply $\mathrm{P}\ne\mathrm{NC}^1$, but this necessary condition is not a security proof. The interface contains DQC1 and supports weighted infinite-temperature observables; a proposed spin-prediction application does not establish a separation. Multiple arithmetic calls and general postprocessing remain outside the simulation theorem, while many clean-state applications, including most quantum chemistry, are unavailable asymptotically.

Little sign of AI-generated textStrong assumptionsOverclaims

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
10 pages, 5 pages of Appendix, 2 Figures, and 1 Table. Comments welcome!

Related papers

Delegation of quantum computation using a random oracle

Jiayu Zhang

Quantum computers can make use of quantum mechanics to achieve surprising speed-ups relative to classical computers, for important computational problems. Quantum computers are gradually coming into reality, but the current implementations are big, require advanced facilities and could only be used as a cloud service in the foreseeable future. If we want to run quantum computations on secret data, we require protocols for performing computations on a remote quantum server without leaking information to the server about the data being processed. In the quantum computation delegation problem, a client with limited quantum computing resources wishes to evaluate a secret quantum circuit C with the help of a more powerful remote server. The protocol is deemed secure if the server learns nothing about either the circuit or the input state even if the server acts maliciously. This problem, first raised in (Childs05), has become fundamental to the study of quantum cryptography, because of its practical importance, and because it provides a testbed for new techniques that can be later applied to related problems (for example, quantum computation verification [FK12]). Prior to our work, known protocols for this problem were either proven secure against all possible attackers, without any complexity-theoretic assumptions, but highly inefficient for the client; or efficient but proven secure only based on strong cryptographic assumptions. This second category of protocols is only secure if there exist public-key encryption schemes that are secure against quantum attacks and that have special additional properties. Very few candidates for such schemes are known, and their security remains poorly understood. We seek to understand how efficient quantum delegation schemes can be designed based on weaker or different complexity assumptions. We study how the availability of symmetric-key primitives, modeled by a random oracle, changes the complexity of quantum computation delegation. In contrast with public-key primitives, there are many candidates for symmetric-key schemes secure against quantum attacks. We aim for protocols that require as little quantum computing as possible for the client. Specifically, we ask that the client only needs to run a quantum circuit of size independent of the circuit being evaluated. Known unconditionally-secure protocols all require the client to evaluate quantum circuits of size at least linear in the original circuit size. In this thesis, we construct two new quantum computation delegation protocols. The first protocol is for the delegation of a large special circuit family, which we call "C+P circuits". This scheme is non-interactive, and can be used to delegate a very important quantum algorithm---Shor's algorithm (Shor,97) for integer factorization and discrete logarithms. Our second protocol requires interaction---one quantum message from the client to the server, and many rounds of classical computation---but supports arbitrary quantum computation (instead of a specific circuit family), and is thus much more general than the first protocol in both theory and practice. Both protocols are proved secure in the quantum random oracle model. The protocols can be instantiated in practice with a symmetric-key cipher such as AES.

•Jan 1, 2022•0 citations

Classical simulatability of the one clean qubit model

Tomoyuki Morimae, Takeshi Koshiba

Deterministic quantum computation with one quantum bit (DQC1), or the one clean qubit model, [E. Knill and R. Laflamme, Phys. Rev. Lett. {\bf81}, 5672 (1998)] is a model of quantum computing where the input is the tensor product of a single pure qubit and many completely-mixed states, and only the single qubit is measured at the end of the computation. In spite of its naive appearance, the DQC1 model can efficiently solve some problems for which no classical efficient algorithms are known, and therefore it has been conjectured that the DQC1 model is more powerful than classical computing (under the assumption of $\mbox{BPP}\subsetneq\mbox{BQP}$). Here we show that the output probability distribution of the DQC1 model cannot be classically efficiently approximated (exactly within a polynomial bit length or in the fully polynomial randomized approximation scheme (FPRAS) with at most a constant error) unless $\mbox{BQP}\subseteq\mbox{BPP}$.

arXiv:1405.6840•May 27, 2014•2 citations

Beyond NISQ Assumptions: One-time Memory in the Classically Accessible Random-Oracle Model

Boyang Chen, Tianren Liu, Luojian Wei

Quantum information enables many cryptographic primitives that are impossible in the classical world. A line of works has developed cryptographic protocol under the assumption that quantum adversaries are restricted to noisy intermediate-scale quantum (NISQ) computing power, enabling strong one-time functionalities. But the advent of early fault-tolerant quantum computers eras will allow deeper logical quantum circuits, calling into questions the applicability of these NISQ-based assumptions. In this work, we adapt the classically accessible random oracle model (CAROM) as in [BDF+11] and [AK22], in which adversaries are only allowed to classically query the random oracle. The restriction is well motivated for NISQ quantum adversaries and may remain plausible in the presence of early fault-tolerant quantum computers. Then, we show that an efficient simulation-secure one-time memory (OTM) is possible under CAROM. Our protocol uses only BB84 states and has quadratic communication: for a $λ$-bit message and integer-valued parameters $n=n(λ)$ and $\ell=\ell(λ)$, the construction uses $n\ell$ qubits and $(n+2)λ$ classical bits, and for any quantum adversary with at most $2^{\ell/2-1}-1$ classical queries to the random oracle, its simulation advantage is at most $(n+3)\left(\frac{3}{4}\right)^n.$ Hence exponentially small simulation advantage in $n$.

quant-pharXiv:2609.32603•Sep 29, 2026•0 citations

Hardness of Classically Simulating the One-Clean-Qubit Model

Tomoyuki Morimae, Keisuke Fujii, Joseph F. Fitzsimons

Deterministic quantum computation with one quantum bit (DQC1) [E. Knill and R. Laflamme, Phys. Rev. Lett. 81, 5672 (1998)] is a model of quantum computing where the input is restricted to containing a single qubit in a pure state and has all other qubits in a completely mixed state. Only the single pure qubit is measured at the end of the computation. While it is known that DQC1 can efficiently solve several problems for which no known classical efficient algorithms exist, the question of whether DQC1 is really more powerful than classical computation remains open. In this Letter, we introduce a slightly modified version of DQC1, which we call DQC1(k), where k output qubits are measured, and show that DQC1(k) cannot be classically efficiently simulated for any k≥3 unless the polynomial hierarchy collapses at the third level.

arXiv:1312.2496•Physical Review Letters•Apr 2, 2014•108 citations

The Space Just Above One Clean Qubit

Dale Jacobs, Saeed Adel Mehraban

Consider the model of computation where we start with two halves of a $2n$-qubit maximally entangled state. We get to apply a universal quantum computation on one half, measure both halves at the end, and perform classical postprocessing. This model, which we call $\frac12$BQP, was defined in STOC 2017 [ABKM17] to capture the power of permutational computations on special input states. As observed in [ABKM17], this model can be viewed as a natural generalization of the one-clean-qubit model (DQC1) where we learn the content of a high entropy input state only after the computation is completed. An interesting open question is to characterize the power of this model, which seems to sit nontrivially between DQC1 and BQP. In this paper, we show that despite its limitations, this model can carry out many well-known quantum computations that are candidates for exponential speed-up over classical computations (and possibly DQC1). In particular, $\frac12$BQP can simulate Instantaneous Quantum Polynomial Time (IQP) and solve the Deutsch-Jozsa problem, Bernstein-Vazirani problem, Simon's problem, and period finding. As a consequence, $\frac12$BQP also solves Order Finding and Factoring outside of the oracle setting. Furthermore, $\frac12$BQP can solve Forrelation and the corresponding oracle problem given by Raz and Tal [RT22] to separate BQP and PH. We also study limitations of $\frac12$BQP and show that similarly to DQC1, $\frac12$BQP cannot distinguish between unitaries which are close in trace distance, then give an oracle separating $\frac12$BQP and BQP. Due to this limitation, $\frac12$BQP cannot obtain the quadratic speedup for unstructured search given by Grover's algorithm [Gro96]. We conjecture that $\frac12$BQP cannot solve $3$-Forrelation.

arXiv:2410.08051•Oct 10, 2024•11 citations

Succinct blind Quantum computation using a random oracle

Jiayu Zhang

In the universal blind quantum computation problem, a client wants to make use of a single quantum server to evaluate C|0⟩ where C is an arbitrary quantum circuit while keeping C secret. The client’s goal is to use as few resources as possible. This problem, first raised by Broadbent, Fitzsimons and Kashefi [FOCS 2009], has become fundamental to the study of quantum cryptography, not only because of its own importance, but also because it provides a testbed for new techniques that can be later applied to related problems (for example, quantum computation verification). Known protocols on this problem are mainly either information-theoretically (IT) secure or based on trapdoor assumptions (public key encryptions). In this paper we study how the availability of symmetric-key primitives, modeled by a random oracle, changes the complexity of universal blind quantum computation. We give a new universal blind quantum computation protocol. Similar to previous works on IT-secure protocols (for example, Broadbent-Fitzsimons-Kashefi), our protocol can be divided into two phases. In the first phase the client prepares some quantum gadgets with relatively simple quantum gates and sends them to the server, and in the second phase the client is entirely classical — it does not even need quantum storage. Crucially, the protocol’s first phase is succinct, that is, its complexity is independent of the circuit size. Given the security parameter κ, its complexity is only a fixed polynomial of κ, and can be used to evaluate any circuit (or several circuits) of size up to a subexponential of κ. In contrast, known schemes either require the client to perform quantum computations that scale with the size of the circuit, or require trapdoor assumptions.

arXiv:2004.12621•ACM Symposium on Theory of Computing (STOC)•Jun 15, 2021•15 citations
Publicly accessible fault-tolerant quantum computers must preserve scientific utility while limiting cryptanalytic power. We propose the restricted model of computation, 1/2BQP_1: a quantum server provides random computational-basis inputs, revealed only after execution, and one designated output bit. This interface permits arbitrary circuits and system sizes. We conjecture that a classical client making polynomially many adaptive requests cannot efficiently factor RSA moduli. The model subsumes the well-known one-clean-qubit model DQC1, retaining its applications to infinite-temperature multi-time correlations, out-of-time-order correlators, and suitably normalized partition functions. We propose a candidate for separating 1/2BQP_1 from DQC1 based on testing classical predictions of quantum spin dynamics. We argue security in two steps. First, we show that one-bit readout makes the quantum stages of standard factoring algorithms classically simulable, even when their preparation and postprocessing are redesigned within a broad family around a single arithmetic-oracle call. This covers the quantum stages of Shor, Ekera-Hastad, Kitaev, and Regev factoring constructions. Second, we examine the workaround of coherently implementing rational reconstruction to release a factor bit. Random inputs obstruct this straightforward attack: known techniques accommodate logarithmic-depth classical circuits, but rational reconstruction has resisted such parallelization for decades. Therefore, we show that security of our protocol is built on not only classical hardness for factoring, but also conjectured circuit lowerbounds for rational reconstruction.
The paper asks whether public quantum computers can retain useful simulation capabilities while restricting cryptanalytic computation through input/output rules rather than circuit inspection. It proposes 1/2BQP11/2\mathrm{BQP}_11/2BQP1​: the client specifies an arbitrary polynomial-size unitary, the server runs it on a uniformly random computational-basis input, measures one designated output qubit, and reveals the input only afterward. The central conjecture is that polynomially many adaptive requests, with polynomial-time classical processing, cannot efficiently factor semiprimes.
The main simulation theorem covers circuits A=NOπVA=\mathcal N O_\pi VA=NOπ​V, with computationally tractable preparation VVV, one efficiently invertible arithmetic permutation OπO_\piOπ​, and normalizer postprocessing N\mathcal NN. On every basis input, the released bit is classically sampleable to total-variation error ε\varepsilonε in time polynomial in circuit description length and 1/ε1/\varepsilon1/ε; this extends to adaptive transcripts. The family includes the quantum stages of Shor, Ekerå–Håstad, Kitaev, and Regev constructions. The proof combines tractable-state expectation estimation with a character expansion of the measured bit whose coefficient weight is O(m)O(m)O(m) for mmm qubits.
A conditional attack shows that rational reconstruction in uniform NC1\mathrm{NC}^1NC1 would enable factoring even with DQC1 access, using Barrington-style reversible evaluation and repeated phase estimation. Thus security would imply P≠NC1\mathrm{P}\ne\mathrm{NC}^1P=NC1, but this necessary condition is not a security proof. The interface contains DQC1 and supports weighted infinite-temperature observables; a proposed spin-prediction application does not establish a separation. Multiple arithmetic calls and general postprocessing remain outside the simulation theorem, while many clean-state applications, including most quantum chemistry, are unavailable asymptotically.
Quantum computers can make use of quantum mechanics to achieve surprising speed-ups relative to classical computers, for important computational problems. Quantum computers are gradually coming into reality, but the current implementations are big, require advanced facilities and could only be used as a cloud service in the foreseeable future. If we want to run quantum computations on secret data, we require protocols for performing computations on a remote quantum server without leaking information to the server about the data being processed. In the quantum computation delegation problem, a client with limited quantum computing resources wishes to evaluate a secret quantum circuit C with the help of a more powerful remote server. The protocol is deemed secure if the server learns nothing about either the circuit or the input state even if the server acts maliciously. This problem, first raised in (Childs05), has become fundamental to the study of quantum cryptography, because of its practical importance, and because it provides a testbed for new techniques that can be later applied to related problems (for example, quantum computation verification [FK12]). Prior to our work, known protocols for this problem were either proven secure against all possible attackers, without any complexity-theoretic assumptions, but highly inefficient for the client; or efficient but proven secure only based on strong cryptographic assumptions. This second category of protocols is only secure if there exist public-key encryption schemes that are secure against quantum attacks and that have special additional properties. Very few candidates for such schemes are known, and their security remains poorly understood. We seek to understand how efficient quantum delegation schemes can be designed based on weaker or different complexity assumptions. We study how the availability of symmetric-key primitives, modeled by a random oracle, changes the complexity of quantum computation delegation. In contrast with public-key primitives, there are many candidates for symmetric-key schemes secure against quantum attacks. We aim for protocols that require as little quantum computing as possible for the client. Specifically, we ask that the client only needs to run a quantum circuit of size independent of the circuit being evaluated. Known unconditionally-secure protocols all require the client to evaluate quantum circuits of size at least linear in the original circuit size. In this thesis, we construct two new quantum computation delegation protocols. The first protocol is for the delegation of a large special circuit family, which we call "C+P circuits". This scheme is non-interactive, and can be used to delegate a very important quantum algorithm---Shor's algorithm (Shor,97) for integer factorization and discrete logarithms. Our second protocol requires interaction---one quantum message from the client to the server, and many rounds of classical computation---but supports arbitrary quantum computation (instead of a specific circuit family), and is thus much more general than the first protocol in both theory and practice. Both protocols are proved secure in the quantum random oracle model. The protocols can be instantiated in practice with a symmetric-key cipher such as AES.
Deterministic quantum computation with one quantum bit (DQC1), or the one clean qubit model, [E. Knill and R. Laflamme, Phys. Rev. Lett. {\bf81}, 5672 (1998)] is a model of quantum computing where the input is the tensor product of a single pure qubit and many completely-mixed states, and only the single qubit is measured at the end of the computation. In spite of its naive appearance, the DQC1 model can efficiently solve some problems for which no classical efficient algorithms are known, and therefore it has been conjectured that the DQC1 model is more powerful than classical computing (under the assumption of $\mbox{BPP}\subsetneq\mbox{BQP}$). Here we show that the output probability distribution of the DQC1 model cannot be classically efficiently approximated (exactly within a polynomial bit length or in the fully polynomial randomized approximation scheme (FPRAS) with at most a constant error) unless $\mbox{BQP}\subseteq\mbox{BPP}$.
Quantum information enables many cryptographic primitives that are impossible in the classical world. A line of works has developed cryptographic protocol under the assumption that quantum adversaries are restricted to noisy intermediate-scale quantum (NISQ) computing power, enabling strong one-time functionalities. But the advent of early fault-tolerant quantum computers eras will allow deeper logical quantum circuits, calling into questions the applicability of these NISQ-based assumptions. In this work, we adapt the classically accessible random oracle model (CAROM) as in [BDF+11] and [AK22], in which adversaries are only allowed to classically query the random oracle. The restriction is well motivated for NISQ quantum adversaries and may remain plausible in the presence of early fault-tolerant quantum computers. Then, we show that an efficient simulation-secure one-time memory (OTM) is possible under CAROM. Our protocol uses only BB84 states and has quadratic communication: for a λλλ-bit message and integer-valued parameters n=n(λ)n=n(λ)n=n(λ) and ℓ=ℓ(λ)\ell=\ell(λ)ℓ=ℓ(λ), the construction uses nℓn\ellnℓ qubits and (n+2)λ(n+2)λ(n+2)λ classical bits, and for any quantum adversary with at most 2ℓ/2−1−12^{\ell/2-1}-12ℓ/2−1−1 classical queries to the random oracle, its simulation advantage is at most (n+3)(34)n.(n+3)\left(\frac{3}{4}\right)^n.(n+3)(43​)n. Hence exponentially small simulation advantage in nnn.
Deterministic quantum computation with one quantum bit (DQC1) [E. Knill and R. Laflamme, Phys. Rev. Lett. 81, 5672 (1998)] is a model of quantum computing where the input is restricted to containing a single qubit in a pure state and has all other qubits in a completely mixed state. Only the single pure qubit is measured at the end of the computation. While it is known that DQC1 can efficiently solve several problems for which no known classical efficient algorithms exist, the question of whether DQC1 is really more powerful than classical computation remains open. In this Letter, we introduce a slightly modified version of DQC1, which we call DQC1(k), where k output qubits are measured, and show that DQC1(k) cannot be classically efficiently simulated for any k≥3 unless the polynomial hierarchy collapses at the third level.
Consider the model of computation where we start with two halves of a 2n2n2n-qubit maximally entangled state. We get to apply a universal quantum computation on one half, measure both halves at the end, and perform classical postprocessing. This model, which we call 12\frac1221​BQP, was defined in STOC 2017 [ABKM17] to capture the power of permutational computations on special input states. As observed in [ABKM17], this model can be viewed as a natural generalization of the one-clean-qubit model (DQC1) where we learn the content of a high entropy input state only after the computation is completed. An interesting open question is to characterize the power of this model, which seems to sit nontrivially between DQC1 and BQP. In this paper, we show that despite its limitations, this model can carry out many well-known quantum computations that are candidates for exponential speed-up over classical computations (and possibly DQC1). In particular, 12\frac1221​BQP can simulate Instantaneous Quantum Polynomial Time (IQP) and solve the Deutsch-Jozsa problem, Bernstein-Vazirani problem, Simon's problem, and period finding. As a consequence, 12\frac1221​BQP also solves Order Finding and Factoring outside of the oracle setting. Furthermore, 12\frac1221​BQP can solve Forrelation and the corresponding oracle problem given by Raz and Tal [RT22] to separate BQP and PH. We also study limitations of 12\frac1221​BQP and show that similarly to DQC1, 12\frac1221​BQP cannot distinguish between unitaries which are close in trace distance, then give an oracle separating 12\frac1221​BQP and BQP. Due to this limitation, 12\frac1221​BQP cannot obtain the quadratic speedup for unstructured search given by Grover's algorithm [Gro96]. We conjecture that 12\frac1221​BQP cannot solve 333-Forrelation.
In the universal blind quantum computation problem, a client wants to make use of a single quantum server to evaluate C|0⟩ where C is an arbitrary quantum circuit while keeping C secret. The client’s goal is to use as few resources as possible. This problem, first raised by Broadbent, Fitzsimons and Kashefi [FOCS 2009], has become fundamental to the study of quantum cryptography, not only because of its own importance, but also because it provides a testbed for new techniques that can be later applied to related problems (for example, quantum computation verification). Known protocols on this problem are mainly either information-theoretically (IT) secure or based on trapdoor assumptions (public key encryptions). In this paper we study how the availability of symmetric-key primitives, modeled by a random oracle, changes the complexity of universal blind quantum computation. We give a new universal blind quantum computation protocol. Similar to previous works on IT-secure protocols (for example, Broadbent-Fitzsimons-Kashefi), our protocol can be divided into two phases. In the first phase the client prepares some quantum gadgets with relatively simple quantum gates and sends them to the server, and in the second phase the client is entirely classical — it does not even need quantum storage. Crucially, the protocol’s first phase is succinct, that is, its complexity is independent of the circuit size. Given the security parameter κ, its complexity is only a fixed polynomial of κ, and can be used to evaluate any circuit (or several circuits) of size up to a subexponential of κ. In contrast, known schemes either require the client to perform quantum computations that scale with the size of the circuit, or require trapdoor assumptions.