Every 2026 trend list says AI agents are coming to freight. Most carriers are asking a more practical question: will any of it actually work on MY freight?

It is the right question. And the honest answer depends less on the AI than on what sits underneath it.

Why bolt-on AI guesses

Generic AI added on top of batch data has a problem: it can only reason from what the operation recorded.

If the pickup was never scanned, the AI does not know the freight moved.
If the POD is a photo of a clipboard, the AI cannot read the exceptions.
If the data arrives as an end-of-day export, the AI is always describing yesterday.

Ask that system a question and it will still give you an answer. It will just be a guess dressed up as one.

The digital twin underneath

Hitrak was built event-driven: every scan, pickup and delivery event updates a live digital twin of the operation as it happens.

That is the data our built-in AI copilot works on - a transport management system (TMS) for freight carriers where the intelligence sits on live order data, not last week's export.

In practice, that looks like:

Ask it to summarise a consignment's history, and it reads the actual event trail.
Ask it to investigate a claim, and it works from what was captured at the dock, in the truck, at delivery.
Ask it what a new driver should be trained on, and it reasons from real run data.

Not a demo layer

The test of operational AI is whether it holds up inside a running operation.

Carriers using Hitrak automate up to 95% of dispatch and cut POD chasing by 90%. The copilot is not a feature bolted on for the sales deck - it is sitting on the same live data the dispatch board runs on.

For New Zealand carriers weighing up AI claims, one question cuts through most of the noise: show me the data the AI reasons from, and show me when it was captured.

Real AI needs real data. That is the whole thesis.