In the paper era, every contract was executed in counterparts — one for each party. A digital contract gives you none: there is only the provider’s live page, changeable at any moment. Doctri restores contract law’s oldest safeguard to the digital age.
The problem is not that terms are long or unread.
It is structural: the text you accepted does not exist on your side.
You accepted a platform’s terms. Where does the text live? On the provider’s server. Who controls it? The provider. How would you show what you consented to? You cannot — you hold no counterpart. In classical contracting, each party keeping a copy was never a formality; it was the balance of power that made silent alteration impossible. In digital contracting that balance has collapsed in one direction: the contract stops being a contract and becomes a unilateral notice that can be rewritten, retroactively, at any time.
Legal traditions that never met each other invented the same mechanism: never leave the document in one party’s hands.
The contract was written twice on a single parchment, then cut apart along a jagged line. Alter your half, and the teeth no longer match. The word itself comes from “indent” — to cut like teeth. Its canon-law cousin: the chirograph.
A debt was notched onto a wooden stick, and the stick split lengthwise. Each side kept a half; no record was valid unless the halves matched. Unilaterally “updating” the record was physically impossible.
Contracts executed in as many originals as there are parties, each keeping its own; from qadi registers to notarial archives, the record is never left to one party’s mercy. The counterpart is a constitutive safeguard of contract itself.
Privacy policies, terms of service, EULAs: the only “copy” is a page on the provider’s server, changeable at will. A balance built over millennia disappeared everywhere at once — silently.
If every civilization converged on the same solution, the value protected is not a local preference — it is a structural precondition of contract itself.
In electronic form, the counterpart is constituted not by a copy kept by each party, but by a record both parties can access and neither can alter alone.
This is not a claim that “new technology demands new law.” It is the opposite: the value the old law protected has been lost in the digital environment, and we are putting it back. An argument of continuity, not rupture — that is where its doctrinal force comes from.
The counterpart is a concept about the plurality of custody. Hand the record to an escrow third party and it still sits in one place; the custodian changes, the imbalance is inherited. Its digital equivalent is, by definition, a distributed ledger.
Our demand is embarrassingly modest — which is exactly why it cannot be refused.
Let providers change their terms whenever, and however, they wish. We ask for no content review, no approval mechanism, no bureaucracy. The entire obligation reads: “when you publish your document, record it.” This is not a compliance burden but a requirement of form — its cost is marginal, and identical for the smallest and the largest provider.
This is the only thing we restrict. A provider objecting to it must say, out loud: “I want my changes to go unseen.” No platform can utter that sentence in public. It is a demand whose refusal is shameful.
The sanction creates no vacuum — there is always a governing text. It is self-executing — no court needs to debate which clause governs; it looks up the last recorded version. And it is proportionate — failing to record yields no advantage and triggers no catastrophe: the provider has simply not made the change.
Major legal orders already reject silent unilateral change. What is missing is not the norm, but the record showing which text changed, and when.
Douglas v. Talk America (9th Cir. 2007): silently posting revised terms binds no one — “parties have no obligation to check the terms on a periodic basis to learn whether they have been changed.”
Harris v. Blockbuster (2009) and In re Zappos (2012): a contract the provider may change at will is illusory — in law, no contract at all.
Douglas v. U.S. Dist. Court ex rel. Talk America, 495 F.3d 1062 (9th Cir. 2007) · Harris v. Blockbuster Inc., 622 F. Supp. 2d 396 (N.D. Tex. 2009) · In re Zappos.com (D. Nev. 2012)The annex to Directive 93/13/EEC lists unilateral-change clauses among presumptively unfair terms. The P2B Regulation mandates advance notice of changes; the DSA requires informing users of significant changes.
And the critical development: eIDAS 2 created the “qualified electronic ledger” — a legal presumption of integrity and chronological order. The EU is standardizing the very instrument this doctrine needs.
Directive 93/13/EEC, Annex 1(j) · Regulation (EU) 2019/1150 · DSA Art. 14 · Regulation (EU) 2024/1183 Art. 45l (implementing standards in progress)The legal order that invented the indenture and the tally stick still polices unilateral change: the Consumer Rights Act 2015 places unilateral-variation clauses on its indicative list of unfair terms.
And the case-law “red hand rule”: an onerous, unusual term binds only if fairly brought to the other party’s attention. An invisible change, by definition, is not.
Consumer Rights Act 2015, Sch. 2 paras. 11 & 13 · Interfoto Picture Library v. Stiletto [1989] QB 433Art. 21: standard terms the adhering party was given no real opportunity to learn are deemed unwritten.
Art. 24: clauses granting the drafter a unilateral power to amend are themselves deemed unwritten. “We may change these terms at any time” is already void under Turkish law.
See also: Consumer Protection Law Art. 5 (unfair terms) · Law No. 5070 (qualified timestamps) · Code of Civil Procedure, rules of evidence.
Then why can platforms still change their terms silently?
Because the fact that would trigger Arts. 21 and 24 cannot be proven: nobody can show which text became what, and when. The record sits with the defendant. The rule stays on paper.
For thirty years UNCITRAL has been building the functional-equivalence series: the electronic equivalent of writing was recognized, then of the signature, then of the original — and most recently, with MLETR, of the transferable record. One link remains unnamed: the functional equivalent of the counterpart. Doctri defines it.
And as shown above, US case law, EU legislation, UK law and the Turkish Code all point the same way. What this project offers a court is not a new doctrine but the evidentiary ground on which existing doctrine can finally operate. The missing exhibit in your pleading comes from here.
Hashes fix content, signatures prevent denial, timestamps prove time. All three are valuable — and all three are attached to a document held by one party. They cannot solve the counterpart problem, because the counterpart is plurality of custody: a record both parties can access and neither can alter alone.
We wrote the specification technology-blind, then eliminated instruments one by one: a single party’s server fails; an escrow third party fails; a permissioned consortium fails — members who agree can rewrite, and the counterpart belongs to a club. When the elimination ends, one structure is left standing: a permissionless, sufficiently distributed ledger. Not ideology — a conclusion forced by the doctrine.
| Function | Required instrument | Blockchain needed? |
|---|---|---|
| Fixing the content | Hash | No |
| Non-repudiation by the provider | Digital signature | No |
| Proof of time | Qualified timestamp | No |
| A counterpart held by no one | Distributed ledger | Yes — only here |
These instruments are layers, not rivals: qualified timestamps can and should be used alongside the chain. But a timestamp proves time; it does not create a counterpart. The only structure that does is a ledger no one monopolizes. The doctrine is written technology-neutral; today exactly one class of infrastructure meets the definition — permissionless distributed ledgers. If something better appears tomorrow, the institution survives and the implementation changes.
The provider’s entire obligation is the first three steps. The user never has to do anything.
The platform publishes the new version of its terms exactly as it always has. No one touches the content.
A cryptographic digest (hash) of the text is computed; the platform signs it with its own key. The record is not an observation — it is a declaration of intent.
The signed digest is written to a permissionless, sufficiently distributed ledger. Batched anchoring (Merkle) brings the cost per document down to cents.
User, court, researcher — anyone can compare the digest of the text they are shown against the record. If it does not match, that amendment is void.
What Creative Commons spread was not a technology; it was a license text. What we spread is likewise a legal instrument: any platform that adopts this clause hands its users their counterpart back.
Only the version of this agreement written to the record system is binding between the parties.
An amendment enters into force [N] days after it is recorded; no amendment applies retroactively.
No unrecorded amendment may be invoked against the user; the last recorded version remains in force.
Which version is in force is determined directly from the record system; the validity of an unrecorded version cannot be made a subject of proof. In disputes over whether the text shown matches the recorded version, or over the regularity of the record itself, the burden of proof rests with the platform.
One precedent is worth a thousand volunteer platforms. One consultation response outlasts a thousand posts. Pick your role: litigate, sign, respond, pilot.
The rules against silent unilateral change sit on paper in every jurisdiction; you are the ones who can set them in motion. A single strategic case in which a silent change is finally documentable turns doctrine into precedent — and the signature under that decision will be yours. Let’s build the evidentiary architecture of the file together.
The demand is one sentence, and everyone understands it: “I want my own copy of what I accepted.” Lead the public petition that puts signatures under that sentence. And at the same time, take a seat at the tables where the rules are being written: eIDAS 45l standards, DSA codes of conduct, national secondary legislation — these processes are in motion right now; let’s be at the table before the texts are settled.
We are not asking you for a new rule; we are offering the infrastructure that makes your existing rules auditable. The most realistic first step is one sentence in public procurement templates: “The contractor shall anchor its user-facing terms in a verifiable record.” Privacy notices, subscription contracts, distance-selling terms — the pilot ground is ready, the cost is marginal.
“The functional equivalent of the counterpart” is a concept the literature has not yet named — and the first to name it wins. The doctrinal article is in preparation; we welcome contribution, critique and co-authorship on comparative law, evidence law and IT-law fronts. For law-clinic students, a ready field: documenting silent changes in the wild.
This is a working draft: the doctrinal article, the model clause and the technical specification are in preparation. Lawyers — case law and doctrine. Engineers — anchoring and verification infrastructure. Platform operators — become an early adopter. Write to us.