The patent and the disclosure

There is a temptation to treat a patent as a pure instrument of exclusion — a fence around an idea, a legal weapon for keeping competitors off your ground. That is what it does, but it is not what it is. A patent is a trade. The inventor gives up something valuable and receives something valuable in return, and the thing given up is secrecy. In exchange for a monopoly that runs for a fixed term and then expires, the inventor must lay the invention open: describe it, in the public record, fully enough that a competent stranger could build it. The exclusive right is the reward. The disclosure is the price. And the deepest thing about the arrangement is that society deliberately chose to pay for innovation in the currency of openness rather than let it be hoarded in the dark.

The bargain is old enough to predate the modern state. The Venetian Statute of 1474 is commonly cited as one of the first systematic patent laws — a formal promise that those who devised new and ingenious contrivances would receive protection, provided the device was reduced to practice and made known to the Republic. A century and a half later, the English Statute of Monopolies of 1624 drew the line that still governs. It swept away the crown's habit of selling monopolies over ordinary trades as instruments of royal favour, while carving out a deliberate exception: a genuine inventor of a genuinely new manufacture could still be granted a limited exclusive right. Monopoly as patronage was abolished; monopoly as the earned reward for bringing something new into the world was preserved. The distinction between the two turns entirely on whether the public gets something in return.

Protection bought with disclosure

What matured out of this over the following centuries was the requirement that gives the bargain its teeth: the specification, or what patent law calls enabling disclosure. It is not enough to gesture at an invention or to claim a result. The applicant must describe the invention in terms clear and complete enough that a person skilled in the relevant art could reproduce it without undue experimentation. The document has to teach. If it does not teach — if it hides the crucial step, or claims more than it actually explains — the protection is not earned, and a patent can fail for that reason alone. The law treats an incomplete disclosure as a broken promise, because the whole justification for the monopoly is the teaching that was supposed to come with it.

Set against that requirement, the alternative comes into focus. A monopoly granted without disclosure is not a patent at all; it is simply rent. It extracts a toll from the public and returns nothing to the common store of knowledge. That is precisely the abuse the Statute of Monopolies was written to end. The genius of the patent bargain is that it refuses to let exclusivity and secrecy be held at the same time. You may have your temporary wall around the invention, but only because you have already handed everyone the blueprint that lets them build past it the day the wall comes down. Protection and disclosure are not in tension in this system; disclosure is the thing that makes the protection legitimate.

A monopoly with disclosure is a reward for teaching the world something. A monopoly without disclosure is a toll booth. The patent bargain was built to tell the two apart.

Forcing “how it works” into the record

Consider what society could have chosen instead, and nearly always had chosen before. The natural way to protect a valuable method is to keep it secret — the guild's guarded technique, the family recipe, the process no one outside the workshop is allowed to watch. Secrecy protects the inventor for as long as the secret holds, which can be forever, but it teaches no one and advances nothing. The field stays exactly as ignorant as it was. The patent system made the opposite wager: that a taught, examinable invention is worth more to a civilization than an eternally guarded one, even at the cost of an exclusive right that eventually lapses. It bought disclosure by pricing it — by making the public record of how the thing works the condition of the reward, rather than a courtesy the inventor might extend.

That the record is public and examinable is not incidental. It is the point. A disclosed invention can be studied, improved upon, designed around, and built freely once the term expires. Every specification is a deposit into a commons that compounds, so that each generation of inventors starts from a higher floor than the last. The monopoly is the visible feature, but the disclosure is the mechanism that actually moves the field forward. Society decided, in effect, that the way to reward the clever was to make them explain themselves — and to keep the explanation where anyone could check it.

A disclosure bargain for the decision

Now turn the lens toward how consequential machine decisions are defended today. The reflex runs the other way. A model produces a score, a denial, a ranking, an eligibility judgment, and when the affected party asks how, the answer is that the method is proprietary — a trade secret, a black box, a competitive edge that cannot be shown. The system claims the rewards of operating a valuable decision engine while disclosing nothing about the basis on which it decides. That is the pre-patent instinct exactly: protection through concealment, reward without teaching. It is the toll booth, not the specification.

The patent bargain suggests a modest but pointed analogy — and I want to keep it modest, because the analogy is about disclosure regimes, not about forcing anyone to publish their source code. The lesson of the patent is not that valuable methods must be given away. It is that protection and disclosure can be traded, and that society may legitimately require the second in exchange for the first. A patent owes a disclosure sufficient to reproduce the invention. The analogous bargain for a consequential automated decision is a disclosure sufficient to examine and contest it: the evidence actually relied on, the rules in force at the time, enough of the basis to test whether the decision followed from them. You may keep your commercial edge. What you may not do, where a decision carries public consequence, is claim the standing of a legitimate authority while withholding everything that would let the decision be checked.

This is the same trade the patent system has run for five centuries, moved from the invention to the decision. There, disclosure is the price of a monopoly and the engine of a compounding commons. Here, examinability is the price of consequence — the thing a Decision Receipt is meant to carry so that provenance and contestability are not favours the system may withhold but obligations it has already met. The patent taught us that secrecy and legitimacy are not the same, and that a society is entitled to ask for the second when it grants the first. A machine that decides about people, and asks to be trusted while it does, is asking for a form of protection. The old bargain is the right one to hold it to: keep your edge, but owe a disclosure sufficient to examine what you have done.

— Dispatches · Summit Cognitive

---