What This Evidence Proves, and What It Does Not
The statutes behind electronic records, the rules a court actually applies, and the six claims we will not make about a Pactbound bundle.
A Pactbound bundle proves three things: that specific content existed in a specific state, that it existed no later than a particular moment, and that a named party acknowledged receiving it. It does not prove that a court will admit it, that a card processor will rule your way, or that an insurer will accept it. Those outcomes belong to the people deciding them, and anyone selling you certainty about them is selling something they cannot deliver.
This post covers both halves, because the second half is the one that matters when it counts.
Electronic records are not second-class
Under the federal E-SIGN Act, 15 U.S.C. § 7001, an electronic record or signature in an interstate transaction may not be denied legal effect, validity, or enforceability solely because it is in electronic form.
Read the verbs carefully, because the difference is load-bearing. The statute says may not be denied solely because. It does not say electronic records are automatically accepted. It removes one specific objection and leaves every other rule of evidence exactly where it was.
The Uniform Electronic Transactions Act carries parallel effect at state level, adopted in 49 states plus the District of Columbia. New York did not adopt UETA; it has its own Electronic Signatures and Records Act instead.
The rules a court actually applies
Getting past "it is electronic" leaves authentication, which is Federal Rule of Evidence 901: the proponent must produce evidence sufficient to support a finding that the item is what they claim it is.
Rule 902 asks whether a record was generated by a process that produces an accurate result, and whether someone qualified will certify it. Photo: Eric Lozaga via Pexels. Pexels License.
Two subsections of Rule 902 matter here:
| Rule | Covers |
|---|---|
| 902(13) | A record generated by an electronic process or system that produces an accurate result, shown by a certification of a qualified person |
| 902(14) | Data copied from an electronic device or file, authenticated by a process of digital identification, shown by a certification of a qualified person |
A bundle is built in the shape those rules evaluate: an electronic record, integrity hashes, an external timestamp, and a certification available at the time of submission.
Note the phrase that appears in both: certification of a qualified person. That is a procedural step someone performs when the record is submitted. It is not a property the file carries around on its own. A bundle is structured for that authentication. It does not self-authenticate by existing, and we do not claim it does.
Mapping the mechanisms to the questions
| A court, insurer, or processor asks | Answered by |
|---|---|
| Did the content change? | SHA-256 hashes and a Merkle root, per NIST FIPS PUB 180-4 |
| When did it exist? | A public-ledger timestamp assigned by consensus, which no party to the dispute controls |
| Who agreed? | One-time-code identity check, recorded e-signature consent, IP address, and timestamp per signature |
Three questions, three independent mechanisms. Failing one does not silently pass the others, which is why the verifier reports them separately.
Chargebacks
Stripe's dispute categories ask, for a "product not received" claim, for documentation showing the service was provided, and separately for whether the customer communicated about the issue before filing.
An acknowledgment is contemporaneous, recorded customer communication. It is evidence of the shape a processor asks for, not a guaranteed outcome. Photo: Unknown via Openverse. CC0 1.0.
An identity-verified acknowledgment is that second thing: contemporaneous, recorded, and submittable. It is evidence of the shape Stripe asks for.
What it is not is a guaranteed win. Stripe does not describe any single evidence type as decisive, and neither do we. The card networks decide disputes, and they decide plenty of them against merchants holding good documentation.
The six claims we do not make
These would all be useful in marketing. None are supportable, so none appear anywhere on this site or in the app.
- That a bundle is guaranteed to be admitted as evidence in any specific court or proceeding.
- That a bundle is guaranteed to win a chargeback on Stripe, Square, Adyen, PayPal, or any other processor.
- That a bundle is guaranteed to be accepted by any errors-and-omissions or professional indemnity carrier.
- That Hedera anchors are qualified electronic timestamps under eIDAS. They are not. A qualified timestamp requires an EU Trust Service Provider, and a public distributed ledger is not one.
- That Pactbound substitutes for advice from an attorney admitted in your jurisdiction.
- That a bundle is a legal opinion, a legal instrument, or a certified record outside of arrangements you have made yourself.
Whether a specific record is admitted in a specific proceeding is the court's call. Anyone promising you that outcome is promising something they do not control. Photo: Unknown via Openverse. CC0 1.0.
Our own acceptable use policy makes that binding in the other direction as well: customers agree not to represent bundles as legally binding contracts or court-admissible evidence where they know that representation to be false. Evidence bundles are documentation tools, not legal opinions.
The limit worth knowing about
The ledger anchor proves what and when. It does not prove who.
No personal data is written on chain, by design. Identity rests on the acknowledgment record inside the bundle: a one-time code sent to the signer's address, explicit consent to sign electronically, plus IP and timestamp. That is ordinary, reasonable evidence of who acted, and it is the same class of evidence mainstream e-signature platforms rely on.
It is not cryptographic attribution. If someone tells you their product delivers court-grade proof of identity from an email link, ask which key signed and where the private half lives.
Why publish the limits
Because the alternative is worse for you.
Evidence that does exactly what it claims is more useful under questioning than evidence wrapped in guarantees. Photo: KATRIN BOLOVTSOVA via Pexels. Pexels License.
Evidence that is oversold fails at the moment it is relied on, which is the moment you cannot afford it to. A record that does exactly what it claims, with the boundaries written down in advance, is more useful in a dispute than one wrapped in guarantees that evaporate under questioning.
Everything above traces to a named primary source, consulted directly rather than recalled. If you find a claim on this site that a primary source does not support, tell us and we will correct it.