Most cryptography patents pick a lane: a homomorphic-encryption patent, a key-exchange patent, a signature patent. US11394537B2, “Homomorphic encryption with quantum key distribution encapsulation,” granted to JPMorgan Chase Bank on July 19, 2022, is notable for stacking two frontiers at once — and the combination is the claim, not either piece alone.

The two ingredients solve different problems. Homomorphic encryption (CPC H04L 9/008) lets a party compute on encrypted data without decrypting it. Quantum key distribution (CPC H04L 9/0858) is a physics-based method of agreeing on a shared key whose security rests on quantum mechanics — any eavesdropper necessarily disturbs the quantum states and is detected. The patent claims using QKD to handle the key side of a homomorphic workflow.

“Systems and methods for processing and transmission of encrypted data are provided.”— U.S. Patent No. 11,394,537 source

Reading the independent claim makes the structure concrete. Claim 1 is a method, “implemented by at least one processor,” with three operations in sequence: encrypting a first data set; encapsulating that encrypted data set in a protective layer; and transmitting the encapsulated result to a destination. The cryptographic content is in the two “wherein” clauses that pin down how each step is done — the encrypting “is performed by using a homomorphic encryption (HE) technique,” and the encapsulating “is performed by using a quantum key distribution (QKD) encapsulation technique to generate a QKD-protected layer.” So the claim is not abstractly “use HE and QKD together”; it is specifically HE on the data and a QKD-derived layer wrapped around the resulting ciphertext.

The transport model is the other load-bearing detail. The same claim requires that the communication channel “includes a non-quantum channel over which the QKD-encapsulated encrypted first data set is transmitted and a quantum channel over which a quantum key distribution is conducted.” That two-channel split is exactly how QKD works in practice: the quantum channel carries the photonic states from which a shared secret is distilled, and the conventional channel carries the actual payload protected with that secret. The abstract confirms a single-channel variant is also contemplated — “or a single communication channel to conduct both” — but the headline embodiment keeps the quantum key-agreement and the classical data transfer on separate links.

Why pair them? Because they cover each other's exposed flank. Homomorphic encryption protects data in use but still depends on keys that have to be established and protected; QKD offers a key-establishment channel with a very strong, eavesdropper-evident security property. Encapsulating the homomorphic scheme's keys via QKD is, in concept, defense-in-depth across two independent threat models — one rooted in computational hardness (the HE layer) and one rooted in the physics of measurement (the QKD layer). An adversary would need to defeat both, and the assumptions behind them do not share a common point of failure.

The CPC placement reinforces where the examiner saw the center of gravity. H04L 9/0858 is the subclass for quantum-key-distribution key agreement specifically — not generic key exchange — so the grant is anchored in the QKD-protected workflow rather than in homomorphic encryption broadly. That matters for reading the portfolio: this is a key-distribution claim with a homomorphic payload, not a homomorphic-computation claim that happens to mention quantum.

There is a useful asymmetry in how the two layers fail, and it is what makes the stack more than the sum of its parts. The homomorphic layer's security is computational: it holds as long as the underlying lattice or number-theoretic problem stays hard, which is precisely the assumption a future quantum computer threatens. The QKD layer's security is informational and physical: it does not assume the adversary is computationally bounded at all, because detection of eavesdropping comes from the measurement-disturbance property of quantum states rather than from the difficulty of a math problem. Wrapping the HE ciphertext's key material in a QKD-protected layer therefore hedges the HE scheme's one structural weakness — its reliance on a hardness assumption — with a mechanism that does not share that assumption. That is the conceptual argument the claim's two “wherein” clauses encode.

It is worth being clear-eyed about QKD. It requires specialized hardware and links and is not a drop-in replacement for software key exchange; it is a niche, infrastructure-heavy technology that banks and telecoms have piloted rather than broadly deployed. The dependent claims extend the method to encrypting and transmitting additional data sets and to the system and non-transitory-medium forms of the same workflow, which is standard claim-scope hygiene rather than a second invention. The patent is most interesting as a statement about where a frontier financial-cryptography group thought the puck was going, not as evidence of a mass-market product.

Per the desk's rules: issued grant (B2), not an application; method claim combining two techniques, not a shipped offering. The inventor team includes Pistoia and Polychroniadou, names associated with JPMorgan's well-publicized applied-cryptography and quantum research, which is the obvious context.

For the portfolio reader, the takeaway is that the most ambitious financial-cryptography IP of this period is integrative — combining homomorphic encryption, quantum key distribution, and post-quantum schemes into stacked systems. The claim's discipline is instructive: it does not try to own homomorphic encryption or QKD in the abstract, only the specific act of wrapping an HE ciphertext in a QKD-protected layer and shipping it over a two-channel transport. Whether or not any single stack ships, owning that combination is a hedge across multiple possible futures — and a narrow enough hedge to survive prior art on either component alone.