Cargo theft is no longer just a problem of broken seals, stolen trailers and opportunistic theft on the roadside. That world still exists, but it is no longer the main story. The real shift is happening upstream, inside the digital workflows on which logistics depends every day: quoting, carrier onboarding, load boards, document exchange and payment. That is where fraud is now hiding.
The scale of the problem is rising fast. FBI reporting shows that cargo theft losses in the U.S. and Canada reached nearly $725 million in 2025, up 60% year-on-year. Reported incidents rose 18%, while the average value per theft increased 36% to $273,990. Those figures describe a more organized, more digital and more selective form of crime; one in which criminals have moved from physical access to identity exploitation.
What has changed is not simply the volume of crime, but its architecture. Freight fraud has moved upstream, out of the road and into the workflow. Spoofed credentials, fabricated documents, fake carrier identities, compromised inboxes and cloned domains are now central to many theft and diversion schemes. The FBI has warned that threat actors have been infiltrating broker and carrier systems through spoofed emails and fraudulent web links since at least 2024, using that access to post fake listings, impersonate legitimate firms and reroute shipments. The attack surface has expanded from the physical supply chain to the digital identity layer that supports it.
In other words, the most damaging breaches no longer require forcing entry. They simply require logging in.
This is where the industry needs to rethink its assumptions about AI.
Too many conversations still present AI as a pure efficiency tool: faster processing, lower costs, fewer manual checks. That framing is incomplete and, in the context of rising fraud, potentially dangerous. Speed without validation creates scale for whatever is already in the system, and that includes fraud. If AI is used to automate weak processes, or doesn’t take into account enterprise-grade security thinking, it accelerates exposure. Organizations that have digitized fastest without redesigning their controls are, paradoxically, among the most vulnerable. The same capabilities that help legitimate operators move faster with automated approvals, AI-assisted onboarding and rapid document processing can be turned against them if verification is not built into the process from the start.
The real opportunity is to use AI differently: to verify better. That means building systems capable of confirming whether a carrier is real, whether a bank account matches the legal entity, whether a document is consistent with the route and shipment history, and whether a transaction behaves like a genuine business or a manufactured fraud. It means moving from static checks to continuous validation, where AI flags anomalies, scores risk and triggers human review when the data does not add up.
This is where “Know Your Supplier” becomes essential. In financial services, Know Your Customer is a basic compliance requirement because identity risk sits at the heart of the business model. Logistics needs the same discipline. Supplier risk shifts with ownership changes, subcontracting arrangements, cyber exposure, geography, sanctions status and payment behavior. A one-time onboarding check is not sufficient in an environment where identities can be fabricated, and credentials can be cloned at scale. Continuous supplier intelligence with ongoing monitoring, rather than periodic due diligence, is the standard logistics companies need to move toward.
The operational question this raises is: Can your workflows independently verify a legal entity, confirm a bank account, validate a carrier identity, validate supplier quality standards, and detect document inconsistencies before a load is released? For many organizations, the honest answer is no. This is not because the intent is absent, but because compliance has traditionally been treated as a function that sits beside operations rather than as a design principle inside it. That separation is no longer sustainable. If fraud is embedded in workflows, validation must be embedded there too.
This also requires a rethink of what resilience means for logistics businesses. Physical redundancy matters, but it is not sufficient on its own. A supply chain can have capacity, visibility, and speed, yet remain fragile if counterparty data is weak or if onboarding systems cannot reliably distinguish a legitimate vendor from a fabricated one. The boundary between technology risk and operational risk has collapsed. Fraud used to happen at the edge of the supply chain. Now it is inside the workflow, and once that happens, every efficiency gain becomes a potential attack surface unless validation is built in from the start.
The response should be to make digital transformation more rigorous. The companies that will lead the next phase understand a straightforward principle: AI is a force multiplier for organizations that already have discipline. That means building systems where compliance is continuous, data is trusted, and AI is used to strengthen judgment rather than replace it. Compliance and operational intelligence need to work together, not in parallel.
AI will shape the future of logistics. Its value will depend entirely on whether businesses use it to strengthen judgment, not replace it. In logistics, speed has always mattered. In the AI era, trust matters more.
Adrian Smith is CEO and co-founder of Ripple.
