AI and the Freight Forwarder: What’s Really Changing in TMS

Oct 27, 2026 ·

CargoWise has been the default operating system for freight forwarders for years, and it remains the dominant platform in the market. But commentators are increasingly asking whether agentic AI is starting to erode that dominance. This piece separates the genuine shift from the headline noise: where AI is already doing real work inside freight operations, where the TMS itself is still doing the heavy lifting, and what that split means for how forwarders should be investing over the next 12 months.

The CargoWise/AI debate in context

The debate started in the stock market, not on the operations floor. In early 2026, a broad sell-off hit software companies worldwide as investors worried that AI agents from companies like Anthropic and Google could lower the barriers to entry for established platforms. Analysts nicknamed it the “SaaSpocalypse”, and WiseTech was caught up in it alongside almost every other software business.

The question being asked was simple: if AI can write code and run workflows, does a forwarder still need a large, all-in-one TMS?

The views have split along familiar lines:

  • The sceptics point out that AI tools are getting cheaper and more capable, and that some forwarders may build their own AI capabilities rather than buy them from their TMS provider.
  • The defenders, including several equity analysts, argue the fears have gone too far. Macquarie’s view is that CargoWise outputs carry real liability and need to be right every time, which is a very different problem from writing code.
  • WiseTech itself has leaned in rather than pushed back. In February 2026 it announced a plan to cut around 2,000 roles as part of becoming an “AI-led” organisation, and it is building AI agents directly into CargoWise.

Meanwhile, the platform has held its ground with customers. WiseTech’s FY26 results report that customer attrition has stayed below 1%, as it has for the last 14 years, even through a major change to its pricing model. Morningstar describes WiseTech as an order of magnitude larger than its nearest competitors.

So the real question for forwarders isn’t whether CargoWise survives AI. It’s which parts of the job AI is genuinely taking over, and which parts still need a platform underneath.

Where AI adds value today

The AI that is genuinely working in freight today is narrow and practical. It takes over repetitive, high-volume tasks that used to need a person to read, type and chase. The clearest examples:

  • Reading documents. Pulling data from invoices, bills of lading, packing lists and arrival notices is now reliable enough for production use on clean documents, with messy ones flagged for a person to check.
  • Chasing missing information. WiseTech’s Smart Auto-Request Agent watches a live shipment, spots what’s missing, requests it from the right party and brings the answer back into CargoWise.
  • Quoting and rate sheets. AI can read carrier rate sheets and assemble quote responses far faster than a person working through emails and spreadsheets.
  • Drafting customer emails. Routine updates and replies can be drafted automatically, freeing the team to focus on the conversations that need judgement.
  • Suggesting customs codes and flagging exceptions. AI can propose a classification or highlight a shipment that looks wrong, but a person still signs off.

The common thread is that AI is doing the work around the transaction, not replacing the system that holds it. Most of these tools, whether built by WiseTech or by specialist start-ups, read from and write back into the TMS.

WiseTech has set out an ambitious target for this: it says its CargoWise AI capabilities are aimed at up to around 50% labour cost savings for logistics providers. That’s a stated goal rather than a measured result, but it shows where the platform owner believes the value lies.

What still needs a platform and an implementation partner

AI agents are only as good as the system they plug into. The heavy lifting in freight still sits in the TMS, for a few clear reasons:

  • A single source of truth. Shipments, costs, invoices and customs records need to live in one trusted place. An agent that reads a document still needs somewhere reliable to put the result.
  • Compliance carries real consequences. A wrong customs entry can stop a shipment or trigger a penalty. Keep skilled people making the final decision whenever a compliance obligation is being actioned.
  • Connections to the wider industry. Carriers, customs authorities and partners are connected through the platform. These links take years to build and can’t simply be recreated by a new AI tool.
  • Clean, consistent data. AI depends on good data. If your CargoWise setup has inconsistent codes, workarounds or duplicated records, AI will automate these rather than fix it.

This is where an implementation partner earns its keep. Getting value from AI is less about switching on a feature and more about the groundwork:

  • Mapping your real processes before deciding what to automate
  • Cleaning up configuration and data so agents have something reliable to work with
  • Deciding where a person must stay in the loop, and setting clear approval points
  • Connecting specialist AI tools to CargoWise without creating new data silos
  • Training teams so they trust and actually use the new ways of working

Industry research backs this up. The biggest blockers to agentic AI in logistics tend to be legacy systems, fragmented integrations, poor data quality and unclear process ownership. None of those are solved by the AI itself.

What to do about it now

The forwarders who get ahead over the next 12 months won’t be the ones who buy the most AI. They’ll be the ones who prepare their platform to use it well. Five practical steps:

  1. Get your CargoWise house in order. Review your configuration, data quality and workarounds. A clean setup is the foundation every AI capability depends on.
  2. Start with one well-defined use case. Pick a high-volume, repetitive task such as document entry or missing-information requests. Prove the value, then expand.
  3. Look at what’s already in CargoWise first. Before buying a separate AI tool, check what’s available through your current setup and the platform’s own AI agents.
  4. Measure outcomes, not features. Track time saved per shipment, error rates and turnaround times, so you know what’s actually working.
  5. Keep people in control of the risky decisions. Be clear about where AI can act alone and where a person must approve, especially for customs and finance.

AI is changing how freight work gets done, but it’s building on the platform, not replacing it. The opportunity is real for forwarders who treat it as an operational change, not a software purchase.

Want to understand where AI could make the biggest difference in your operation? Talk to the Zinnovate team about getting your platform ready.

Sources