The Patent Layer: Tokenization's Newest Market-Structure Problem

Every market structure eventually acquires a legal layer that its architects did not design for. Equities got Reg NMS. Dark pools got Form ATS-N. Tokenized securities are now getting theirs — and it is arriving not as a regulation but as a patent portfolio. On June 15, 2026, tZERO put the tokenization industry on notice that the architecture of compliant digital securities may already be claimed.

The dispute that followed — a cease-and-desist against Securitize, answered a week later by a declaratory-judgment complaint in Delaware — is at its earliest procedural stage. No court has ruled on anything. But the questions it raises reach far beyond two companies: whether the standard architecture of a compliance-enforcing security token is proprietary, whether open standards protect their implementers, and how a young industry prices patent risk it has largely ignored. This report maps the asserted patents, the verified case record, the legal framework, and the freedom-to-operate playbook — and closes with the market-structure convergence that puts this dispute squarely inside Sapinover's coverage universe.

01The Tokenized Universe

Sizing tokenized markets is an exercise in scenario comparison, not consensus reading. The three most-cited forecasts differ in scope, methodology, and horizon — they are not comparable numbers, and treating them as a single range misreads all three. What they agree on is direction: a multi-trillion-dollar asset base migrating onto programmable settlement rails within a decade.

The regulatory scaffolding is arriving in parallel. On January 28, 2026, SEC staff across three divisions published a Statement on Tokenized Securities that draws a structural distinction the market had been blurring: issuer-tokenized instruments, where the issuer itself places the security on a ledger, versus third-party-tokenized models, where an intermediary wraps exposure to an underlying security it holds. The distinction matters here because the patent families examined below claim mechanics — registries, transfer controls, order authorization — that appear in both models.

02IP as an Infrastructure Layer

The tokenization stack is conventionally drawn in six layers: legal rights (what the token represents), identity and compliance (who may hold it), issuance and lifecycle (how it is created, serviced, and retired), custody, cash, and settlement (where it rests and how it moves against payment), trading and liquidity (how it changes hands), and interoperability and governance (how systems and rules connect). The tZERO campaign adds a seventh layer that sits across all of them: intellectual property.

A patent is an exclusionary right — the right to exclude others from practicing the claimed invention, not a right to practice it yourself. That asymmetry is what makes a patent layer structurally different from a regulatory layer. A regulation tells you what you must do; a patent tells you what you may not do without a license, and a single patent family can cut across identity, transfer control, custody, order authorization, and routing simultaneously. An architecture that is fully compliant with every securities regulation can still, in principle, read on someone's claims.

03The tZERO–Securitize Case

The record below is limited to verified public events. It deliberately excludes anything not yet confirmed on the docket or in primary announcements.

One piece of market context belongs in the record, stated factually and without causal attribution: Securitize began trading on NYSE as SECZ on July 2, 2026, following its combination with Cantor Equity Partners II, and declined roughly 40% in its first week of trading. Public evidence does not isolate any single cause for that move, and this report does not attempt to.

Because early-stage patent disputes are routinely over-read, it is worth being precise about what the current record does and does not establish.

04Claim Language Controls

The single most common analytical error in reading patent disputes is treating a patent's title or its marketing description as its scope. Neither is. Under 35 U.S.C. § 271, direct infringement generally requires practicing every element of at least one claim — the all-elements rule. A system that implements nine of ten limitations of a claim does not infringe that claim. A system that does something that sounds like the patent's title, through a different mechanism, may not infringe at all.

This is also why the differentiation questions attached to each patent below matter more than the patent summaries themselves. Each question targets a specific claim limitation whose absence, if established, defeats infringement of that claim.

05Patent-by-Patent: The Verified Record

Six families dominate the enforcement picture: the two asserted patents, the three families the June 15 announcement identified as under investigation, and the continuation activity that shows the portfolio is still expanding. All bibliographic data below is verified against USPTO and Google Patents records.

ASSERTED · GRANTED JAN 4, 2022 · PRIORITY AUG 10, 2018US 11,216,802 — the compliance-registry patent

The official title is a claim summary in itself: “Self-enforcing security token implementing smart-contract-based compliance rules consulting smart-contract-based global registry of investors.” Claim 1 recites a separate compliance smart contract consulted at transfer time, a smart-contract-based global registry of investors, indicators for Regulation A, Regulation D, and Regulation S status, and handling of frozen accounts. The patent also describes PII-hash commitments and AML/KYC features; whether those appear in claim 1 specifically requires full claim review, so this report attributes them to the patent's description rather than to any particular claim.

ASSERTED · GRANTED JUL 19, 2022 · PRIORITY FEB 9, 2015US 11,394,560 — the order-authorization patent

The '560 is the sleeper in the asserted pair, and its priority date is the reason: February 9, 2015 — materially earlier than the 2018 compliance family and years before ERC-1400 or ERC-3643 existed. The claims center on a common identifier linking an exchange order and two accounts, the same key signing both the asset-transfer transaction and the order transaction, and dual signature verification performed before an exchange order is authorized. Any prior-art program aimed at this family has to search pre-2015 material — a far thinner record than the 2018-era compliance-token landscape.

UNDER INVESTIGATION · GRANTED SEP 24, 2024 · PRIORITY AUG 9, 2018US 12,099,996 — the whitelist-verification patent

Issued from application US 2020/0051072 A1, with a priority date one day earlier than the '802. The claims cover whitelist membership plus private-key-association verification before a transfer of a self-regulating token executes. Where the '802 claims a registry architecture, the '996 reaches the wallet-level gating pattern that many permissioned-token systems use in some form.

UNDER INVESTIGATION · GRANTED JUL 18, 2023 · PRIORITY MAY 1, 2015US 11,704,733 — the multi-asset creation patent (+ US 12,469,079)

Another 2015-priority family. Claim 1 of the '733 covers a signed multi-asset creation request, escrow of the underlying, a second-key mint, and an on-ledger record of the created asset. Redemption, release, and destruction mechanics are covered by the patent beyond claim 1. The family is not static: continuation US 12,469,079 granted November 11, 2025 — eight months before the enforcement campaign went public — which is the clearest signal in the record that the portfolio is under active expansion, with claims being drafted against a market that now exists.

UNDER INVESTIGATION · GRANTED SEP 6, 2022 · PRIORITY MAR 15, 2018US 11,436,673 — the consolidated order book patent

The '673 reaches venue mechanics rather than token mechanics: a consolidated order book, custodial escrow of assets pending execution, order aggregation, external routing, and failure-handling redirection when an external venue cannot fill. For trading venues and broker-dealers, this is the family whose claim language sits closest to how digital-securities markets actually route flow.

PORTFOLIO · FOUR ADDITIONAL VERIFIED GRANTSThe wider portfolio

Beyond the six families above, the verified portfolio includes US 10,937,083 (granted March 2, 2021; priority July 3, 2017), covering decentralized fair ordering and matching of transactions; US 10,831,902 (granted November 10, 2020), covering time-centric Merkle hash tree data verification, known as TOME; US 10,552,829 (granted February 4, 2020; priority May 26, 2015), covering obfuscation of intent in transactions; and US 12,348,621 (granted July 1, 2025; priority February 26, 2020), covering secret splitting with verification metadata. None of these has been asserted, but together they show the portfolio's breadth: matching, data integrity, transaction privacy, and key management, in addition to compliance and custody.

06Live Patent Inventory

The table below is served from Sapinover's verified patent inventory and reflects the ten records examined in this report.

07Standards and Prior Art

The chronology is the uncomfortable part for the open-standards community. ERC-3643, the permissioned-token standard whose architecture — on-chain identity registries, compliance contracts consulted at transfer time — overlaps conceptually with the '802 family, was proposed on July 9, 2021 and now carries Final status. The '802's priority date is August 10, 2018. ERC-1400's originating GitHub issue #1411 was opened September 13, 2018 — about a month after the '802 priority date. And the '560 and '733 families reach back to February and May 2015, years before either standard existed.

Two caveats keep that chronology from being a conclusion. First, conceptual overlap is not infringement — whether any ERC-3643 implementation reads on the '802 claims is an implementation-specific, element-by-element question, not a standards-level one. Second, priority dates define where prior art must be found, not whether it exists. A serious prior-art program against these families must search the 2014-to-mid-2018 record: early security-token pilots, public repositories, transfer-agent systems, and pre-standard permissioned-token experiments. That record is thinner than the post-2018 landscape, but it is not empty.

The standards arrived after the filings. That is precisely why the prior-art search has to look where the standards are not.

08Legal Vulnerabilities

An asserted patent is a set of litigation questions, and the defense side of this dispute has a conventional sequence to work through. Non-infringement and claim construction come first: most patent cases are won or lost on how the court construes disputed claim terms, and the differentiation questions in Section 05 are, in effect, a preview of the construction fights.

Second is subject-matter eligibility. Alice Corp. v. CLS Bank, 573 U.S. 208 (2014), established the two-step framework under Section 101: is the claim directed to an abstract idea, and if so, does it contain an inventive concept beyond the idea itself? Financial-computer-implementation patents have fared poorly under Alice, and claims reciting compliance checks and registries on generic ledger infrastructure will face that framework squarely. It is a credible defense theory — not an automatic outcome. Courts have also upheld software claims that recite specific technical architectures, and the asserted claims here are more architectural than the typical Alice casualty.

Third are the classical validity grounds: anticipation under Section 102, obviousness under Section 103, and definiteness, written description, and enablement under Section 112. The 2015-priority families narrow the 102/103 search window; the 2018 families face a richer prior-art landscape.

Finally, remedies. eBay v. MercExchange, 547 U.S. 388 (2006), ended automatic injunctions in patent cases: a prevailing patentee must satisfy the traditional four-factor equity test, and the public-interest factor carries real weight where the accused systems are regulated financial infrastructure with live issuances and investor holdings on them. Even a clean win on liability does not mechanically translate into shutting an accused platform down.

09What It Means for Market Participants

The immediate effect of an enforcement campaign is not any court outcome — it is the diligence burden it imposes on everyone in the blast radius while the outcome is pending. That burden distributes unevenly.

Four structural consequences follow. An IP risk premium enters tokenization economics: legal clearance becomes a line item alongside compliance and audit. Architecture becomes legal strategy: the choice between an on-chain global registry and off-chain eligibility checks is no longer purely an engineering trade-off. Startups bear a disproportionate burden: a freedom-to-operate study and defensive filing program is a rounding error for an incumbent and a funding round for a seed-stage team. And all three of those push toward consolidation — toward platforms large enough to carry the legal overhead, license, or countersue.

Some commentators have compared broad early patent positioning in emerging infrastructure to early internet domain-name speculation; the comparison is imperfect — patents are examined rights granted in exchange for disclosure — and the market will judge through litigation outcomes and licensing behavior.

10The Freedom-to-Operate Playbook

For teams building in this space, the response to an enforcement campaign is a process, not a panic. The sequence below is the standard freedom-to-operate discipline, ordered:

1. Architecture record. Produce an accurate, current description of how your system actually works — registries, keys, escrow, routing — before anyone maps claims against it.
2. Family maps. Chart each relevant patent family: parents, continuations, pending applications, and foreign counterparts. A design-around against a granted claim can be recaptured by a pending continuation.
3. Element-by-element claim charts. Map every limitation of every independent claim against the architecture record. This is the core artifact; everything else supports it.
4. Prosecution history review. What the applicant surrendered to get the claims allowed constrains how broadly those claims can be construed.
5. Prior-art search. Target the window the priority dates define — for these families, 2014 to mid-2018 security-token pilots, repositories, and transfer-agent systems.
6. Design alternatives. For each claim element your system arguably practices, identify and cost the architecture that does not practice it.
7. Contracts. Review vendor and customer agreements for IP indemnities, and negotiate them where absent.
8. Defensive portfolio. File on your own innovations; defensive publication is the cheaper alternative where filing is not warranted.
9. Escalation triggers. Define in advance what event — a letter, a filing, a licensing approach — moves the matter from monitoring to counsel.

11Scenarios

Five outcomes bound the space. Note the last one: a confidential settlement — statistically the most common endpoint for patent litigation — is also the outcome that resolves the least for everyone who is not a party to it.

12The Market-Structure Sidebar

There is a detail in this dispute that connects it directly to the venue universe Sapinover tracks every day. tZERO is not only a patent holder — tZERO ATS, LLC is itself a FINRA-registered, SEC-reporting alternative trading system for digital securities, with a public Form ATS-N on EDGAR (CIK 1193821), and in 2026 it received FINRA approval to trade tokenized corporate debt. It is a non-NMS digital-securities venue: structurally a different animal from the NMS-equity overnight ATSs — BlueOcean, Bruce, Moon — whose tape we process daily, but a Form ATS-N filer all the same, with the same disclosure obligations and the same regulatory perimeter.

That matters because the two threads of market evolution this site covers are converging on the same end state. Extended-hours NMS trading is stretching the regular-session equity market toward continuous operation — NSCC 24x5 clearing live, SIP extended hours approved for December 2026. Tokenized settlement is building continuous, programmable settlement from the other direction. The end state both point at is the same: markets that do not close, settling on rails that execute rules as code. The overnight ATS tape is the empirical preview of the demand side; the tokenization stack is the infrastructure bid for the supply side.

The patent layer now sits across both. The families examined in this report claim transfer control, order authorization, custody lifecycle, and routing — mechanics that any continuous, programmable market needs, whether it grows out of an NMS venue extending its hours or a tokenization platform extending its asset coverage. Whoever holds enforceable claims on those mechanics holds a position in both futures. That is why a patent dispute between two digital-securities firms belongs in a market-structure publication: it is not a sideshow to the 24/7 transition. It is a fight over who owns the toll booths on the road to it.