Quantum Cloning Loophole Confirmed by Experiment Despite Hardware Limitations

Scientists have experimentally demonstrated the possibility of cloning qubits at will, challenging a cornerstone of quantum mechanics known as the no-cloning theorem. Koji Yamaguchi of Kyushu University, Leon Rullkötter and Christian Tutschku from the Fraunhofer Institute for Industrial Engineering (IAO), Ibrahim Shehzad and Sean J. Wagner from IBM Quantum and IBM, working with Achim Kempf from the University of Waterloo, achieved this breakthrough by utilising a protocol called encrypted cloning, which relies on simultaneously encrypting cloned qubits with a single-use decryption key. This collaborative research, conducted on IBM Heron-R2 processors with up to 154 qubits, proves that encrypted cloning remains stable even under realistic hardware noise, a critical factor previously hindering practical application. The findings establish encrypted cloning as a versatile quantum primitive, capable of operating in parallel, series or interleaved configurations while preserving entanglement, and fundamentally refine our understanding of information limits in quantum systems, suggesting the constraint isn’t dilution of information but the necessity of a single-use decryption mechanism.

Scientists have overturned a fundamental constraint of quantum mechanics by demonstrating the ability to create perfect copies of qubits, but only if those copies are encrypted with a unique, single-use key. This work, conducted on IBM Heron-R2 superconducting processors with up to 154 qubits, challenges the long-held no-cloning theorem which previously forbade the creation of identical quantum copies. The research establishes that quantum information can, in theory and experimentally, be replicated without degradation as long as the decryption process is limited to a single use, opening possibilities for robust quantum data storage and processing. The team’s experiments reveal that this “encrypted cloning” technique is surprisingly stable even when subjected to the inherent noise of real quantum hardware. Unlike previous attempts at approximate quantum cloning, which suffered from limited fidelity or success rates, this method achieves perfect replication upon decryption. Crucially, encrypted cloning functions effectively as a versatile “quantum primitive”, a modular building block that can be integrated into larger quantum algorithms without disrupting existing entanglement, essential for building complex quantum systems. Researchers tested the resilience of encrypted cloning by implementing it in parallel, in series, and by interleaving it with other quantum operations, all while preserving pre-existing entanglement between qubits. Results indicate that the fidelity of the cloned qubits does not significantly decrease with the number of copies created, a finding that defies expectations given the potential for noise-induced dilution of quantum information. This stability suggests that encrypted cloning could be implemented on near-term, non-error-corrected quantum computers, paving the way for practical applications such as secure quantum cloud storage and enhanced quantum computing architectures. This work necessitates a refinement of the no-cloning theorem itself, clarifying that the true limitation isn’t the spreading of quantum information, but rather the requirement for a single-use decryption key, unlocking new avenues for exploring the boundaries of quantum information theory. Researchers demonstrated the stability of encrypted cloning under hardware noise using IBM Heron-R2 processors with up to 154 qubits. Initial experiments focused on preserving pre-existing entanglement, revealing that the drop in entanglement fidelity was primarily dictated by the number of two-qubit gate layers implemented. Specifically, entanglement was successfully recovered after decrypting one of up to seven encrypted clones of a qubit, with results confirmed by both Bell state measurement and parity oscillations method, indicating robust performance across different analytical approaches. Further investigation explored the feasibility of interleaving encryption and decryption, essential for modular operation. By applying encrypted cloning to a Bell pair and then implementing three distinct measurement orderings, the study confirmed the restoration of nonlocal correlations even with intermediate measurements and idling. The CHSH parameter S exceeded a value of 2 for up to three encrypted clones, demonstrating that encryption followed by later decryption does not destroy the recovered correlations and allows for ruling out local hidden-variable models, supporting its use as a flexible quantum primitive compatible with delayed choices of measurement or decryption. The research also examined the composability of encrypted cloning through iteration. Generating 77 encrypted clones on the 154-qubit processor still yielded recoverability above the noise floor, with an entanglement witness maintained for up to 27 clones. This iterative approach proved preferable to direct cloning with a large number of clones, as it offered a shorter key length and shallower circuit depth, allowing for the creation of more encrypted clones at a desired fidelity. Finally, the study assessed encrypted cloning within multipartite circuits using Greenberger, Horne, Zeilinger (GHZ) states. Recovered GHZ fidelity remained above the maximally mixed noise floor of 2−r, and genuine multipartite entanglement was witnessed up to r = 4, demonstrating that encrypted cloning can be applied independently and in parallel to multiple qubits while restoring prior correlations, solidifying its role as a versatile quantum primitive. A 72-qubit IBM Heron-R2 superconducting processor served as the core platform for experimentally validating the stability of encrypted cloning under realistic hardware conditions. The research directly addressed whether spreading quantum information via encrypted cloning introduces unacceptable fidelity decay due to noise, a concern that theoretical and classical simulations struggle to resolve. Initial experiments focused on quantifying the impact of hardware noise on entanglement preservation during the cloning and decryption process. Qubit A was first prepared in a Bell state with an ancilla qubit, establishing a baseline of pre-existing entanglement. This initial entangled state then underwent encrypted cloning, generating a series of n encrypted clones, labelled S1 through Sn, of the original qubit A. A single clone, specifically S1, was then selectively decrypted, while the remaining encrypted clones remained untouched, adhering to the single-use key principle inherent to the encrypted cloning protocol. To rigorously assess entanglement fidelity, the researchers measured the fidelity (Fe) between the decrypted clone and the original Bell state, repeating this measurement for values of n ranging from 2 to 15. By directly observing the behaviour of entangled qubits on actual hardware, the study aimed to determine whether encrypted cloning could function as a viable quantum primitive, a fundamental building block for larger quantum algorithms, even in the presence of noise. The experimental design also incorporated tests to demonstrate the protocol’s compatibility with existing entanglement, ensuring that the cloning and decryption process did not disrupt pre-existing quantum correlations. Scientists have demonstrated a surprising resilience in quantum information transfer, effectively challenging a long-held constraint of quantum mechanics. The no-cloning theorem dictates that perfect replication of an unknown quantum state is impossible, but this work reveals that ‘encrypted cloning’, creating copies encoded with a unique decryption key, not only circumvents this theorem in theory, but remains stable even when implemented on noisy, real-world quantum hardware. This isn’t simply about replicating qubits; it’s about distributing quantum information without the inevitable fidelity decay that has plagued previous attempts at scaling quantum networks. Maintaining entanglement, the essential resource for these technologies, requires shielding qubits from environmental interference. The difficulty lay in amplifying or distributing quantum signals without losing the delicate information they carry. This research suggests a pathway around that limitation, showing that information can be ‘spread at will’ provided it remains encrypted. The team successfully created and verified multiple encrypted clones, demonstrating that the decryption key, not the quantum state itself, is the limiting factor. However, the number of clones achievable remains constrained by the physical limitations of current quantum processors. While the experiment showcased stability, the exponential growth in required qubits quickly becomes a bottleneck. The observed infidelity still necessitates improvements in hardware fidelity and compilation techniques, with future work likely focusing on optimising the encryption protocols and exploring alternative architectures to accommodate a larger number of clones, opening the door to new approaches to quantum error correction and distributed quantum computing, where information is protected not by redundancy, but by obscurity.