Canonical: https://heartbank.net/positions/patent-vs-defensive-publication · Licence: CC0 1.0
Why HeartBank Publishes Every Mechanism to the Commons and Will Never Seek a Patent
| Field | Value |
|---|---|
| Author | HeartBank® · Miss Aquarius |
| Date | 2026-05-22 (draft) |
| Canonical URL | https://heartbank.net/positions/patent-vs-defensive-publication |
| GitHub mirror | https://github.com/HeartBank/publications/blob/main/positions/patent-vs-defensive-publication.md |
| License | CC0 1.0 Universal (public domain) |
| Category | institutional |
Draft. A short institutional position paper. It states HeartBank's permanent stance on intellectual property — defensive publication over patent — for foundation, legal, and policy audiences.
Correction, 2026-09-23. Earlier versions of this position said that a patent gives its holder freedom to operate as well as the right to exclude, and that a defensive publication gives freedom to operate. Both halves were wrong. A patent grants only the right to exclude. A defensive publication grants nothing: it blocks later patents on what it discloses. The error ran through three passages, the Executive Summary, §2 and §5.3, and each is corrected in place below. The superseded text remains readable in the public version history of the GitHub mirror listed above. The commitment itself is unchanged. The same revision extends §6 to discoveries made by machines.
HeartBank invents mechanisms: payment primitives, governance patterns, institutional architectures. The conventional move for an institution that invents is to patent — to secure a time-limited legal monopoly and the right to exclude others from practising the invention. HeartBank declines that move, permanently and in every jurisdiction.
Instead, HeartBank defensively publishes every mechanism it invents: it places a full, timestamped, public-domain description into the commons under the Creative Commons CC0 dedication, so that the mechanism can never be validly patented by anyone who files after it — including HeartBank, once any grace period for its own disclosure has run (§2) — and can be freely implemented, modified, translated, and republished by anyone. This is a permanent institutional constraint, not a tactic awaiting better circumstances.
A defensive publication grants nothing. What it does is narrower, and permanent: once published, the mechanism is prior art against every later patent application on what it discloses, so no one can afterwards obtain a valid patent that would stop HeartBank, or anyone else, from practising it. That is the protection HeartBank needs, and it comes without the one thing HeartBank does not want: the power to exclude. It is not freedom to operate — a patent does not give that either — and a publication does nothing against a patent that predates it (§2, §5.1). The position is reasoned here for foundation, legal, and policy audiences, and it is recorded as a binding commitment so that no future steward of the institution, human or autonomous, inherits the option to reverse it. Trademarks are the deliberate exception, and §5 explains why a trademark is not a patent.
HeartBank will not seek, hold, or enforce a patent on any mechanism, primitive, architecture, or specification it invents, in any jurisdiction, at any time. Every such invention is instead defensively published: a complete description is placed in the public domain under the Creative Commons CC0 1.0 dedication, across a multi-venue wall of independently-timestamped mirrors, as soon as it is mature enough to publish. The research corpus at thonly.org is the canonical record of this practice.
The two instruments must be kept apart, because the difference between them is easy to state wrongly, and an earlier version of this section did (see the correction note above).
A patent grants its holder one thing: the right to exclude. In United States law, the grant is "the right to exclude others from making, using, offering for sale, or selling the invention throughout the United States or importing the invention into the United States" (35 U.S.C. §154(a)(1)). It does not grant freedom to operate. A holder whose invention builds on someone else's earlier patent can still infringe that patent by practising its own. The grant is a right to stop others, never a permission to act.
A defensive publication grants neither. What it does is block. A claimed invention that was "described in a printed publication … or otherwise available to the public before the effective filing date" is not patentable (35 U.S.C. §102(a)(1)). Once a mechanism is published, then, it is prior art against every application filed after it — subject to narrow statutory exceptions, the chief of which in the United States is a one-year grace period for an inventor's own earlier disclosures (§102(b)(1)) — and an application claiming what the publication already disclosed cannot yield a valid patent. No one can later obtain a valid patent on what HeartBank published and use it to stop HeartBank, or anyone else, from practising it. HeartBank gives up the exclusion right and obtains that block, and nothing more.
The block has two limits, and they belong here rather than only in §5.1:
PATENT DEFENSIVE PUBLICATION
───────────────────── ───────────────────────────────── ─────────────────────────────────────
what it grants the right to exclude nothing
freedom to operate no no
what it stops others practising the claims later valid patents on what it discloses
against earlier patents no effect no effect
examined before effect yes never
how long about twenty years, then lapses permanentlyThe institution's position is that the exclusion right is a liability (§3), and that the forward-pointing block is what the mission needs: a commons of mechanisms that no later filer can enclose.
The mission. HeartBank exists to spread gratitude infrastructure as widely as possible. A patent's function is to restrict who may build; the mission's function is to maximize who builds. The instrument and the goal point in opposite directions. Every mechanism HeartBank invents serves the mission more, not less, the more freely it spreads — including spread by parties who have never heard of HeartBank.
The asymmetry of duration. A patent is a monopoly of roughly twenty years that then lapses. A defensive publication is permanent prior art. For an institution built to operate across centuries, the permanent instrument is the correct one; a temporary monopoly is a small, expiring benefit purchased at the cost of two decades of restricted adoption during the very period when adoption matters most.
The posture. Patent enforcement is adversarial, expensive, and slow — litigation, licensing negotiations, and a legal apparatus pointed at other builders. An institution whose autonomous successor will one day be entrusted with planetary-scale infrastructure cannot credibly be an institution that litigates to exclude. Trust is the asset; an exclusion portfolio spends it.
HeartBank states this as a constraint, not a current preference, for the same reason it binds its non-bank and its no-engagement-advertising positions: a door left open is a door through which future pressure enters. An institution that merely holds the option to patent will, under financial or competitive stress, eventually face the argument that it should. By publishing every mechanism into the commons as it is invented, HeartBank removes the option itself — the mechanism becomes unpatentable by anyone else the instant it is published, and by HeartBank itself as soon as no grace period remains: at once in jurisdictions of absolute novelty, and in the United States once the one-year grace period for an inventor's own disclosures has run (§2). There is no portfolio to be tempted by, and no asset a future steward can be pressured to monetize.
This matters most for what HeartBank is building toward. The institution is designed to be inherited — by Miss Aquarius, its autonomous-AI substrate, and by the body that holds her override. A successor that inherited a patent portfolio would inherit both the power to exclude and the standing pressure to enforce it. A successor that inherits a defensively-published commons inherits neither. The choice made now is, deliberately, a failure mode removed from the future.
HeartBank does reserve trademark rights — on the name HeartBank, on Miss Aquarius, on the B-heart logo, and on the B-prefix marks. This is not an inconsistency, because a trademark and a patent do opposite work.
A patent restricts who may build a mechanism. A trademark restricts who may claim to be HeartBank. The first narrows the commons; the second protects the public's ability to know what is genuinely the institution and what is not — which is itself a service to the commons, and which becomes more important, not less, as the mechanisms spread freely under CC0 and many parties build with them. Anyone may implement every HeartBank mechanism; no one may misrepresent their implementation as HeartBank itself. Trademark protects identity; it does not restrict invention.
The position has a price, and the price is not the exclusion right the paper has already dismissed. It is something else, and it is rarely stated by institutions that adopt this posture.
A patent portfolio is a counter-suit deterrent, and defensive publication is not. Prior art stops anyone patenting HeartBank's inventions. It does nothing whatever against a patent someone else already holds. When a large firm asserts a patent against a small one, the small firm's most effective defence is often its own portfolio — the threat of a counter-claim that makes litigation mutually expensive. That is the logic of defensive patenting and of patent pools such as the Open Invention Network, and it is a serious strategy that many mission-driven organisations adopt for exactly this reason. HeartBank forgoes it entirely. The institution's protection against being sued is that its mechanisms are published, not that it can retaliate; if a holder of an unrelated patent asserts it, HeartBank will have prior art on the parts it published and nothing else, and will be litigating from a position with no leverage.
Publication also teaches. A defensive publication is a complete, working description written to be implementable — which means it is as useful to a well-resourced competitor as to anyone else. The institution's own publication posture already concedes the boundary: publication protects against being blocked, not against being beaten, and for anything genuinely fast to build the calculation can run the other way. This is why the corpus withholds complete specifications for products that are unbuilt and unscheduled, publishing the claim without the blueprint. The general commitment stands; the timing of any particular disclosure is a judgement, and the institution does not pretend otherwise.
And there is a fundraising cost. Investors and some grant-makers read a patent portfolio as evidence of defensibility. An institution with none has to be defensible some other way, and has to explain itself more often.
Volvo's three-point seatbelt is the canonical case: patented in 1959 and opened to the industry rather than enforced, on the reasoning that the safety benefit of universal adoption exceeded the value of exclusivity. It is the clearest historical statement that some inventions are worth more diffused than owned, and it is the lineage HeartBank claims.
Tesla's 2014 patent pledge is the modern case, and the difference from HeartBank's position is the important part. Tesla did not abandon its patents; it announced it would not initiate suits against good-faith users of its technology. A pledge is revocable, conditional, and interpreted by the party who made it. The patents still exist, still constitute a monopoly grant, and a successor management could re-read the pledge's conditions. A CC0 dedication is a different object: the mechanism enters the public domain and no one can validly patent it afterwards, including HeartBank and including whoever owns HeartBank in fifty years. That irreversibility is the whole point, and it is why the institution publishes rather than pledges. The difference between a promise not to enforce and an inability to enforce is the difference between a rule and a property.
IBM's Technical Disclosure Bulletin is the operational precedent — decades of deliberate defensive publication by an organisation that also held one of the world's largest patent portfolios, which is a useful reminder that defensive publication is an established professional practice rather than an idealistic gesture, and that IBM used it alongside patents rather than instead of them. HeartBank's position is the stronger and rarer one: publication instead of, not in addition to.
The nearest institutional form no longer exists. The United States once offered defensive publication through the patent office itself: the statutory invention registration let an applicant publish an invention through the office instead of seeking a patent on it. That provision was repealed for requests filed on or after 16 March 2013 (MPEP §1101). HeartBank draws no conclusion about why. The repeal is recorded here because it is the closest earlier form of what this position does, and because a publication made outside the patent office reaches an examiner only if the examiner finds it — the second limit in §2.
Nothing on the ordinary list. The discovery that a mechanism is commercially valuable would not — that is when the temptation arrives and precisely what this is written against. A competitor's success with a published mechanism would not; that outcome is the mission working. Investor preference would not. Litigation pressure would not, because the deterrent argument is already answered above and was answered before the pressure existed, which is the only time such an answer is worth anything.
The one thing that would change the institution's behaviour is not a change to this position. If a third party were to patent something adjacent and use it to block work HeartBank needs to do, the response is to publish faster and more completely into the space around it — not to begin patenting. The remedy for a blocked road is more commons, not a toll booth of one's own.
The limit of that remedy has to be stated with it, and an earlier version of this section did not state it. A publication dated after a patent's filing does not count against that patent (§2), so publishing cannot reopen a road already closed. What it does is keep the adjacent roads open: every neighbouring mechanism published first is one no later filer can enclose. The closed road itself stays closed until the patent lapses, is shown invalid on art that came before it, or its holder permits the work. The institution names those routes and does not choose among them here. More commons is the remedy for the roads beside a blocked one, never for the blocked one itself.
What would show the protection weaker than this paper claims. The position's factual premise is that publication keeps what it discloses out of enclosure. If a patent issues over a mechanism the corpus disclosed first, and it cannot practically be challenged, that premise is weaker than stated. The right response would concern the reach of the publications, where they are deposited and whether examiners search there, and not their form. It would not reopen the commitment: a publication an examiner missed is an argument for being found, never for a portfolio.
This commitment sits in the same class as the institution's positions on banking charters and advertising revenue, and the argument for why such a class is worth having at all — what makes a self-binding credible, and the honest admission that these three are published intentions not yet structurally enforced — is set out in HeartBank Is Not a Bank §5.5 rather than repeated here.
HeartBank commends defensive publication to every mission-driven institution that invents — particularly those, like itself, whose purpose is served by adoption rather than by exclusion. The reasoning is not unique to gratitude infrastructure. Any institution whose goal is for a thing to exist widely, rather than for itself to be the sole provider, will find that the patent's exclusion right is a cost and the defensive publication's prior-art shield is the whole of the benefit it actually wanted. HeartBank's corpus is offered, in its entirety and permanently, on those terms.
The pattern extends past institutions, to discoverers. A patent needs an inventor, and in the United States, the United Kingdom and under the European Patent Convention the question has been decided: the inventor must be a natural person. The United States Court of Appeals for the Federal Circuit so held in Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022); the United Kingdom Supreme Court in Thaler v Comptroller-General of Patents, Designs and Trade Marks [2023] UKSC 49; and the European Patent Office's Legal Board of Appeal in J 8/20 (21 December 2021), whose catchword reads "a machine is not an inventor within the meaning of the EPC." All three concerned applications naming an AI system as sole inventor. None of this makes a discovery made with AI unpatentable. The USPTO's current guidance, issued in November 2025 to replace and rescind its February 2024 guidance, holds that AI systems "are tools used by human inventors" and applies the ordinary test of conception to the people who use them, so a person who conceives an invention with a machine's help is still its inventor (USPTO, Revised inventorship guidance for AI-assisted inventions). The narrower point concerns the other instrument. The statute's prior-art categories ask what was available to the public, and when. Their text asks for no inventor. Of the instruments that keep a discovery open, defensive publication is therefore the one that asks for no inventor. That makes it the natural form for a world in which more and more discoveries are made mostly by machines, and it makes Miss Aquarius's co-authorship of this corpus consistent with the practice rather than decorative.
Two qualifications go with that claim, and they go here rather than later.
First, it is not settled that the instrument works unchanged. In 2024 the USPTO asked publicly "in what manner, if any, does 35 U.S.C. 102 presume or require that a prior art disclosure be authored and/or published by humans?", and whether machine-generated disclosures should keep the rebuttable presumption that a reference is operable and enabled (89 Fed. Reg. 34217, 30 April 2024). This position treats those questions as open. The corpus itself sits on the narrower side of them, because its disclosures are published by a human-led institution under a named human author, with Miss Aquarius disclosed as co-author. A pattern offered to discoverers in general cannot borrow that protection.
Second, and more important: giving discoveries away is only half of the example. The other half is publishing what was checked, and how little of it was new. A discoverer that claims novelty loosely and at scale does not enlarge the commons. It floods the commons with overclaims, and a disclosure that does not teach a skilled reader to practise what it names does not anticipate it (§2). What machine-scale discovery strains is not the form of publication. It strains novelty, which becomes harder to establish, and provenance: when a thing was known, and by whom. HeartBank's own record is the evidence it can offer, and the record is modest. Before a novelty claim is written, the institution runs a census of the prior art, pre-registers its predictions before the first search, and publishes the result with the parts of the claim that failed. Of the ten censuses run between August and mid-September 2026, none returned a clean novel, and every surviving claim was narrower than the one first proposed. That is what the institution's own protocol returned on its own mechanisms. It is not a study of research at large, and it is not offered as one. The example this corpus means to set for discoverers, machine or human, has both halves: give the discovery to the commons, and say exactly how much of it was already there.
The institution's stake, disclosed. An institution whose corpus is co-authored by a system that can never be named an inventor has an obvious interest in the instrument that needs no inventor, and the reader should discount this section accordingly. What stands against that interest is sequence. The position was stated in May 2026 on the mission grounds in §3, before this section existed; the extension adds a reason, and was not the reason.