In the week of issued U.S. patents dated April 28, 2026, Citibank, N.A. appears twice in the blockchain grant column, alongside Wells Fargo and nChain at the top of a thin field where the single largest bucket was again unassigned records. A granted patent is enforceable coverage of the claimed method, distinct from a published application that merely discloses intent. Citibank's two new grants — and the estate they extend — point at a specific layer: enforcing rules and signatures on public chains.

The notable grant is US12615160B2, “Systems and methods for enforcing cryptographically secure actions in public, non-permissioned blockchains using bifurcated self-executing programs comprising shared digital signature requirements.” The granted method wraps a first smart contract inside a second one, where the outer program enforces a digital-signature requirement that the inner program does not carry on its own. The claim states the structure directly:

Systems and methods for a bifurcated self-executing program that wraps a first self-executing program (e.g., a first smart contract) on a blockchain within a second self-executing program (e.g., a second smart contract), in which the second self-executing program enforces the digital signature requirement.— Systems and methods for enforcing cryptographically secure actions in public, non-permissioned blockchains using bifurcated self-executing programs comprising shared digital signature requirements, US12615160B2

In plainer terms, it is a way to add an institutional control — a required signature — on top of a smart contract running on a permissionless chain like Ethereum, without rewriting the underlying contract. The second grant from the same week, US12615221B2 (“Systems and methods for managing resources across global or cloud networks”), covers receiving a request to process a blockchain action and executing it under conditions where network resources may go offline. Both grants are about operating on a chain under institutional constraints, not about issuing a token or running an exchange.

A recent, concentrated estate

The two grants sit on top of an issued blockchain estate that the records show has grown quickly. Citibank's blockchain grant count rose from 7 in 2021 to 32 in 2022, 67 in 2023, and 111 in 2024, with the leading CPC class being H04L 9/50 (29 grants) followed by the security classes G06F 21/577 and G06F 21/552. That is a recent build-out, not a legacy holding, and it skews toward security and on-chain program control.

More recent grants extend the same theme. US12645805B1 (“Testing functionality of and generating on-chain programs,” issued June 2, 2026) claims generating and testing on-chain program code against a test blockchain — tooling for building the very smart contracts the bifurcated-program grant governs. US12652172B2 (“Separate encryption of different parameters of an access token”) covers encrypting individual token parameters for different recipients, and US12651078B2 (“Monitoring and controlling communications between autonomous agents”) records agent communications via distributed-ledger technology for audit. The pattern across the estate is institutional control and auditability layered onto chains and ledgers.

Inside the bifurcated program

The claims describe the wrapping mechanism in operational terms and name the trait that makes it enforceable on a public chain. Claim 1 specifies "a layer one blockchain solution featuring bifurcated self-executing programs that are compiled together," and defines the bifurcated program as "a combination of a plurality of individual self-executing programs that cannot be altered once the bifurcated self-executing program is deployed." Immutability after deployment is the load-bearing property: because the wrapper is fused to the inner contract at compile time, the institutional control cannot be stripped off on-chain. On receiving a request to perform a blockchain action, the system first "process[es] the first request with a first self-executing program," "determin[es] whether the first origination corresponds to an origination requirement," and only "in response to determining that the first origination corresponds to the origination requirement" hands the request to "a second self-executing program" to actually perform the action. The outer program is a gatekeeper that checks where a request came from before the inner program runs.

The dependent claims enumerate what "origination" can mean and what else the wrapper can demand. Origination can be determined from "an entity that issued the first request" (claim 3), "a vetted entity" used to set the requirement (claim 4), "an address that issued the first request" (claim 5), or "a unique identifier" (claim 6). Beyond origin, claim 7 adds "a notification requirement" that must notify and receive confirmation before proceeding, and claims 8 through 15 let the program test a "processing characteristic" against requirements covering a "security certificate," a "request validator," a "request validation procedure," a "request testing protocol," a "supported protocol," a "digital asset type," and "synthetic asset support." Claim 16 spells out the build step — "receiving a first uncompiled version," "receiving a second uncompiled version," and "compiling" them together — confirming that the two programs are joined before deployment. The claim set is, in effect, a menu of compliance conditions a regulated institution can bind to a contract running on a permissionless chain.

What the coverage maps to

For a general reader, the business translation is that Citibank's enforceable blockchain coverage is positioned where a regulated bank would need it — adding signature requirements and controls to programs running on public chains, generating and testing on-chain code, and keeping auditable records — rather than at the consumer-token layer. That contrasts with the protocol-mechanics footprint of an assignee like nChain in the same weekly column, and it is consistent with a bank documenting how it would operate on permissionless infrastructure under its own compliance constraints.

The thinness of the field is worth stating: the week of April 28, 2026 produced fewer than 80 blockchain-tagged grants in total, most without a named assignee, so two grants to one bank is a meaningful share of a small set. And the standard caveat holds — a grant establishes that a claim issued, not how broadly it reads or whether it is in production. What the records do show is the direction of Citibank's accumulating coverage: control and enforcement on public chains, with the two April grants squarely within that estate.