In the week of issued U.S. patents dated May 5, 2026, nChain Licensing AG appears three times in the blockchain grant column — among the most of any named assignee that week, in a field where the largest single bucket was unassigned records. A granted patent is different in kind from a published application: it is issued, enforceable coverage of the claimed method for the life of the patent. Reading nChain's three new grants alongside the company's broader issued estate shows where that estate is concentrated, and it is not at the wallet.

The clearest of the three is US12621154B2, “Devices and methods for consolidating serialized tokens.” The granted method takes two or more tokens whose denominations sum to a larger one and merges them into a single new token with a fresh serial number, then deactivates the originals — the on-chain equivalent of trading four quarters for a dollar. Its companion in the same week, US12619982B2 (“Single-use tokens”), covers issuing tokens whose validity is conditioned on a specific blockchain output remaining unspent. Both are about the accounting machinery of tokens — how they are created, denominated, and retired — classified across the G06Q 20/38 payment-architecture group.

The third grant, US12621363B2 (“Layered network”), is lower-level still. It claims a connection protocol for a node network arranged in ordered layers — a core layer, a second layer, and outer layers — where every node must connect to at least one node of a preceding layer and every outer node must also reach a core node. The granted claim ties those core nodes directly to the chain:

A computer-implemented method for connecting to a layered network. The layered network comprises a plurality of nodes arranged in an ordered set of layers.— Layered network, US12621363B2

That is a grant about network topology — how blockchain nodes are wired together — rather than about moving value. Taken together, the three cover token accounting, token validity, and node connectivity: the plumbing beneath a chain, not the user-facing service on top of it.

Where the issued estate sits

The three new grants are consistent with a large existing portfolio. nChain's issued blockchain estate runs to more than 440 patents in the records, and the class distribution is telling: 289 of them carry H04L 9/50 (the blockchain-specific cryptography subclass), with heavy secondary concentrations in H04L 9/3247 (digital signatures), H04L 9/085 (secret sharing), and the G06Q 20 payment-architecture classes. The annual issue counts show the estate is recent and growing — 60 grants in 2021, 103 in 2023, 146 in 2024 — rather than a legacy holding.

Other recent grants fill in the same map. US12651259B2 (“Multi-party blockchain address scheme,” issued June 9, 2026) claims a way to build a locking script around a chain of hashed public keys so that spending requires a target public key, a signature, and an intermediate hash — multi-party control over an address. US12647282B2 (“Generating digital signature shares”) covers threshold-signature construction across a group of participants, and US12640948B2 (“Blockchain transactions including portions of code in different languages for complex validation”) covers running non-native code to feed a value back into a locking script. The throughline across the estate is key construction, conditional unlocking, and transaction-script mechanics.

How the layered network is wired

The "Layered network" grant reads, in its claims, as a topology specification for a blockchain network rather than a generic graph. Claim 1 ties the core layer directly to the chain: "each core node is a respective blockchain node of a blockchain network configured to operate as a blockchain mining node and/or a blockchain storage node," and the connection protocol "requires that each node must connect to at least one node of a preceding layer; and each outer node must also connect to at least one core node." The effect is that every peripheral node, however many layers out it sits, retains a direct path to a mining-or-storage core — a deliberate constraint on how a light client reaches the chain's authoritative nodes.

The dependent claims describe what those connections carry, and the detail is blockchain-specific. A second or outer node is "configured to transmit a blockchain transaction to one, some or all of the core nodes" (claim 9), to "request confirmation … that a blockchain transaction has been accepted in a pool of transactions that have been validated according to a set of consensus rules" (claim 10), to "request … a Merkle proof of a transaction that has been mined in a block" (claim 11), and to "request one or more block headers" (claim 12) — the standard toolkit of simplified-payment-verification clients. Claim 13 lets outer nodes "identify a respective mining node of the core layer based on a respective mining identifier included in a generation transaction of a block mined by that respective mining node." Identity is handled with certificates: nodes are "associated with a respective digital certificate" (claim 14), the protocol can require that "a given node must only connect to nodes that are associated with a respective digital certificate" (claim 15), and core nodes can "issue a respective digital certificate" to the second nodes beneath them (claim 16). Taken together, the claims describe an ordered, certificate-gated mesh whose periphery talks to the chain through Merkle proofs and block headers — protocol plumbing, not a user product.

What the coverage maps to

For a general reader, the business translation is that nChain's enforceable coverage is positioned at the protocol layer of a Bitcoin-style chain — the rules for how tokens are denominated and retired, how addresses and signatures are built from multiple parties, and how nodes connect — rather than at the level of an exchange, a card, or a consumer app. That is a different footprint from the bank and network assignees that also appear in the same weekly grant column, whose blockchain grants more often attach to payment authorization and fraud workflows. It is worth stating plainly that blockchain grant volume overall is thin: the week of May 5, 2026 produced fewer than 100 blockchain-tagged grants total, and most carried no assignee, so three issued patents to one company is a visible share of a small field.

The standard caveats hold. A grant establishes that a claim issued; it does not, on its own, establish how broadly the claim reads or whether it is being practiced. What the records do establish is the shape of the estate: nChain continues to accumulate issued coverage on the mechanics of tokens, keys, and node networks, and its three grants in the first week of May 2026 are squarely within that pattern.