Beyond Maximal Entanglement: Exact Resources for Multiparty Encrypted Quantum Cloning

Abstract

Encrypted quantum cloning distributes an unknown $k$-qubit state among $m$ encrypted clones so that no individual clone reveals the input, yet the state can be recovered from any one clone together with a common quantum key. We ask the inverse question: for a fixed encoding architecture, which multipartite pure states can serve as exact resources for this task? For $m\ge2$, we completely characterize the pure resources compatible with a sector-wise two-Pauli encoder, with necessity holding for arbitrary completely positive trace-preserving (CPTP) recovery maps. For even $m$, exact recovery requires maximal entanglement across the signal--noise cut. For odd $m$, less entanglement can suffice, provided that the surviving signal correlations have the structure selected by the encoder. We further show, without fixing the encoder, that exact recovery from every authorized subsystem already implies perfect concealment of each individual signal and requires at least $(m-1)k$ ebits of signal--noise entanglement. For graph states, the resource classification reduces to an exact condition on the kernel of the signal--noise cut matrix, leading to binary certification and constructive Clifford recovery. We identify rank-deficient graph resources that attain the architecture-independent entanglement bound and prove that the entire Dicke family, including $W$ states, is excluded for every sector-wise two-Pauli encoder. Our results show that encrypted recovery depends not only on how much entanglement a resource contains, but also on how its correlations are organized relative to the encoder.