When Documents Get Cheap, Proof Becomes the Product
Trade verification has always tested documents for plausibility. When plausible documents cost nothing to produce, plausibility stops carrying information.
An Assumption Trade Was Built On
Documentary trade rests on a proposition nobody states because it has been true for four hundred years: a convincing document is expensive to produce. Printing, letterheads, stamps, consistent typography, plausible commercial language, an understanding of what a real bill of lading looks like — each was a barrier, and together they meant that a document which looked right was probabilistically likely to be right. Verification practice was built on top of that economics. Examiners check internal consistency, format, arithmetic and appearance, because historically those were correlated with legitimacy.
Generative tools attack the correlation rather than the checks. When producing a professional-looking commercial invoice, packing list or certificate becomes close to costless, appearance stops carrying information about legitimacy. The checks still run and still catch careless work; what they no longer do is separate the diligent from the fabricated, because the fabricated is no longer careless.
This is not primarily a fraud story, and framing it as one leads to the wrong response. The serious consequence falls on honest traders, whose defence has always rested implicitly on the plausibility of the documents they were handed by counterparties.
The Legal Position Is Unforgiving About Where Error Lands
Trade law does not distribute the cost of a convincing false document toward the party that made it. It concentrates the cost on the party that lodged it.
Article 15(2) of the Union Customs Code makes the person lodging a declaration responsible for the accuracy and completeness of the information given, for the authenticity, accuracy and validity of any supporting document, and for compliance with the obligations of the procedure. The provision carries no knowledge requirement. An importer who receives a well-made false certificate from a supplier, believes it, and lodges on the strength of it has still lodged inaccurate information — and the responsibility sits where the Regulation puts it.
India's structure differs in a way that makes the record even more important. Section 114AA of the Customs Act, 1962 reaches a person who knowingly or intentionally makes, signs or uses a declaration false in any material particular. The mental-state threshold is genuine protection for an honest trader — but only a trader who can demonstrate the honesty. And demonstration is documentary: what was checked, against what, when, and by whom. A trader with a complete record of verification has a defence. A trader with the same good faith and no record has an assertion.
The same logic reaches the intermediary. Regulation 10(e) of CBLR 2018 requires a customs broker to exercise due diligence to ascertain the correctness of information imparted to a client, and Regulation 10(n) requires verification of the client's identity and functioning at the declared address using reliable, independent, authentic documents, data or information. Both are conduct standards. Conduct is proven contemporaneously or not at all.
What Retains Value When Artefacts Do Not
If the document cannot be trusted on its face, the question becomes what can. The answer is structural rather than technological: the properties that survive are the ones a generated artefact cannot manufacture, because they refer to events outside the artefact.
Origin. Not what the document says, but where the datum entered the record — which system, which party, at what point in the transaction. A fabricated invoice can assert any value; it cannot retroactively insert itself into a purchase order raised months earlier, a payment instruction, or a carrier's booking record.
Grounding. Which rule was applied, at the version in force on that date, and what it required. This is checkable independently of the document and of whoever supplied it.
Attestation. A named person who examined the set and signed, at a recorded time, with a record of what they changed. Attribution has a cost that generation does not — the signer is exposed if the signature is wrong, and that exposure is exactly what makes it informative.
Consistency across independent records. A single document can be produced to order. A consistent story across an invoice, a transport document, a bank instruction, a customs declaration and a carrier's own manifest is a considerably harder thing to fabricate, because each element is held by a different party with different incentives.
The Response That Does Not Work
The intuitive response — better detection of generated documents — is worth doing and cannot be the strategy. Detection is an arms race between two systems of the same kind, and the defender has to win continuously while the attacker needs to win once per shipment. Any detection technique that becomes standard becomes a training target.
The structural response is different in kind: stop asking whether the document looks real, and start asking whether the fact it asserts is corroborated by records that were not produced for this purpose. That reframing does not require winning a detection race, because it does not test the artefact at all. It tests the chain.
Where This Leads
The shift is already visible in the direction regulation is travelling. The newest documentary regimes ask for the fact behind the certificate — the plot coordinates, the installation-level emissions, the audited origin — rather than for the certificate itself. Regulators are, in effect, moving from artefact-verification to provenance-verification for their own reasons, and the same move is what protects a trader against convincing counterparty documents.
The commercial implication is that the durable asset in trade compliance is no longer the file of documents; it is the record of what was verified, against which rule, on what evidence, signed by whom. Documents will keep being produced, faster and more cheaply than ever. Proof will not, and proof is what a customs authority, a bank, an insurer and an auditor are each actually asking for.
This is why TradeWatch builds from the shipment's own documents rather than from external signals, why every check carries the instrument that governs it, why unresolved evidence is recorded as unresolved rather than assumed away, and why a named reviewer-of-record signs. Kanan Labs prepares a readiness packet. It does not file Shipping Bills and holds no customs credentials — your licensed CHA files. Kanan Labs prepares claim-admissible evidence. It does not advise on, select, or bind insurance — your IRDAI-licensed broker does.