AI in Freight Forwarding – What's Realistic and What's Not

There's a lot of talk about artificial intelligence. Sometimes too much. Between the headlines and the daily reality of a freight forwarding company, there's often a world of difference. As someone who has been working intensively with AI for more than three years and uses it in his own logistics companies, I want to paint an honest picture here.

People Remain at the Center

First things first: AI should never perform tasks completely on its own. A checking glance from an employee is almost always necessary – and important. This applies to an automatically generated invoice just as much as to an AI-generated email to a customer.

Many people are put off when they hear about so-called hallucinations – AI making things up. That's a legitimate concern. But nowadays, you can prevent this almost 100 percent if you set up the systems correctly. And the remaining percent? An experienced employee catches those with a quick glance. Humans and AI together – that's the combination that works.

The Real Risk Is Doing Nothing

In my assessment, the problem for companies isn't engaging with AI. The problem will be that others are doing it. Intensively.

This leads, on one hand, to competitors having lower costs because they've automated processes that you're still doing manually.

But the far more important point is different: your own customers and service providers will use AI to improve their processes. And that can have direct impacts on your business.

A Concrete Example

Imagine one of your customers sets up AI-powered monitoring. It automatically asks: "Where is my container? Will the delivery time be met?" If no answer comes within five minutes, the AI sends a reminder. And then another one.

Suddenly you have dozens of automated inquiries in your inbox within a very short time – and simply don't have the manpower to handle them all manually. While your competitor, who uses AI themselves, answers these inquiries automatically.

This isn't science fiction. It's technically possible today and will increasingly become reality in the coming months.

What AI Can Realistically Do Today

Works very well: Automatically reading and processing PDF documents. Automatically responding to emails in multiple languages. Automatically querying and preparing tracking data. Creating standard reports and analyses. Checking data quality and detecting errors.

Works well with human oversight: Supporting dispatch and route planning. Price calculation and offer comparisons. Automating customer communication. Preparing customs documents.

Not reliable enough without supervision yet: Conducting complex negotiations. Making independent decisions in exceptional situations. Creative problem-solving for unusual transport requirements.

Honestly...

It's hard to precisely define what AI can and can't do right now. The development is simply too fast. What wasn't possible six months ago works reliably today. I try to stay up to date myself – though honestly, that's not always easy at this pace.

What I can say with certainty, though: the best time to start is now. Not to overhaul everything at once, but to begin with small, concrete applications and gain experience. That's exactly what we help with.

Your contact, Marc