For most of modern shipping history, the chain of custody for high-value cargo has rested on three pillars: a sealed door, a signed bill of lading, and a sampling-based inspection regime at chokepoints. Each was honed in an era when the trade-off looked like this: cargo moves too fast to monitor continuously, so we instead monitor it discretely and accept reasonable doubt in the gaps.
That trade-off is breaking down. The gaps aren't reasonable anymore.
How we got here
The current model is mostly a creature of three documents. The Hague-Visby Rules of 1968 established the bill of lading as the standard evidentiary instrument of who handled what, when. The International Ship and Port Facility Security (ISPS) Code, adopted by the IMO in 2004 in the post-9/11 reset, formalised port-side security regimes. The World Customs Organisation's SAFE Framework and the various Authorised Economic Operator programmes filled in the global customs picture.
What none of these were designed for, and what is now exposing them, is a parallel revolution in how cargo can fail.
Modern containers carry mRNA biologics that lose efficacy if they spend two hours above 8°C. They carry semiconductor wafers that fail silently from a single electrostatic discharge event. They carry food that does not poison anyone but quietly loses a week of shelf life from undocumented temperature deviation. None of those failures are visible to a door seal or a paper waybill.
The shift in regulatory posture mirrors the shift in failure mode. The US Drug Supply Chain Security Act (DSCSA) finished its stabilisation period in November 2024; prescription pharmaceuticals now have to carry verifiable transaction history at unit level across the dispenser network. The FDA's Food Safety Modernization Act Section 204 rule formalises Critical Tracking Events and Key Data Elements for high-risk foods. The EU's NIS2 Directive and CSDDD extend operational resilience and due diligence obligations across the supply chain. In each case, the regulator has moved past "did the carrier sign for it" toward "what actually happened to it".
The insurance signal
Cargo insurers have arrived at the same conclusion via a different path.
The TT Club and the International Union of Marine Insurance have published years of loss analysis showing that the dominant cost of cargo claims is no longer total loss from catastrophic events. It is condition damage, partial loss, and theft sophisticated enough to leave the seal intact. The standard Institute Cargo Clauses, written in their current form in 1982 and revised in 2009, were built around named perils and total-loss arithmetic. They are not naturally suited to "spent four hours at 12°C between Antwerp and Felixstowe".
Parametric cargo insurance has emerged in the last five years as the underwriters' attempt to bridge this gap. Instead of compensating proven losses, parametric structures pay out when an objective parameter, often a temperature or shock threshold, is breached. Lloyd's syndicates and continental reinsurers are quietly writing this paper today. But parametric cover only works if the parameter itself is verifiable and immune to dispute. Self-reported sensor logs from a cargo loader's tablet do not qualify. The data has to be signed at the device, before it is transmitted, by something the insurer trusts.
What "continuous evidence" actually means
The technical shift is from periodic inspection to continuous, signed evidence. Four parts to it, each matters individually.
Multi-modal sensing. Single-sensor instrumentation, the temperature logger you might find clipped to a pallet, was designed to answer one question. Real failure modes need fusion: temperature plus humidity plus shock plus light plus gas plus acoustic plus GNSS plus barometric. A drilling event registers acoustically before it registers anywhere else. A spoofed GPS coordinate is detectable by the inertial sensors that disagree with it. Multi-modal sensing is the difference between detection and inference.
Signing at source. If sensor data is recorded by the device but encrypted only at the cloud, the device-to-cloud link becomes the trust gap. Evidence-grade hardware signs at the device, typically using asymmetric cryptography like Ed25519, before the reading leaves silicon. The signature travels with the data through every intermediate system. Edits become mathematically detectable.
Append-only evidence chain. Each signed reading is hashed into a chain that any subsequent reading depends on. Any attempt to remove or modify a past reading breaks the chain. Some implementations anchor periodic hashes on a public ledger to provide third-party timestamp witness. This is the model regulators are quietly moving toward in pharmaceutical track-and-trace and in food traceability.
Pre-arrival delivery. The evidence packet should reach the receiving party before the cargo does. Customs and port authorities are increasingly equipped to make risk decisions on pre-arrival data. Trusted-trader programmes like C-TPAT in the US and AEO in the EU formalise this. The Green Lane concept, where verified-clean cargo bypasses secondary inspection, is the operational expression of the same logic.
What changes downstream
The implications fan out across the parties involved.
For shippers, the change is from "we monitored it" to "we can prove what we monitored". Disputed claims become rarer because the data is harder to argue with. Insurance premiums fall, slowly but measurably, when underwriters can verify rather than estimate the risk.
For carriers and 3PLs, the change is harder. Continuous evidence makes performance transparent. A carrier with strong cold chain handling sees its premium reflect that. A carrier with sloppy handoffs is now visible to the buyer in a way it was not before. The competitive ground shifts toward operators who can substantiate their service.
For regulators, the change is from auditing samples to interrogating evidence. An FDA investigator reviewing a pharmaceutical recall can ask, of any specific lot, where it sat for how long at what temperature, and get a signed answer. The frequency and granularity of enforcement improves. So does the cost of non-compliance.
For port authorities, the change is the slow rebuild of the inspection regime. A small percentage of containers are physically inspected globally. The choice of which to inspect has historically been a risk model fed by paperwork and intelligence. Continuous evidence feeds that model directly. The capacity saved on verifying compliant cargo gets spent on the cargo that genuinely warrants attention.
The honest version
None of this is finished. The Institute Cargo Clauses have not been rewritten. Most port authorities do not yet have the systems to ingest pre-arrival sensor packets at scale. There are unsolved questions in cross-jurisdiction evidence admissibility. The hardware is good but not free. The regulators are moving but not in step.
What is clear is the direction. Inspection-and-sampling is giving way to continuous-and-signed. The gap between cargo conditions and cargo evidence is closing. The companies and regulators positioning for it now are quietly building a meaningful advantage.
For everyone else, the evidence gap is still there. It just costs more every year to leave it open.