Most of what gets written about AI in freight forwarding is about the network: better route planning, sharper ETA forecasts, smarter pricing. Those gains are real, but they mostly land with carriers and the largest global forwarders, the companies that own enough data to train on. For a typical forwarding operations team, the daily pressure is somewhere else. It is the inbox.
Pre-alerts, carrier updates, arrival notices, pickup orders, invoices and customer questions all arrive as email. Every one of them has to be read, matched to the right shipment and turned into the next action by a person. That is where most of an ops team's hours go, and it is where AI in freight forwarding has the most room to change how the work gets done.
This guide looks at AI from the operator's seat. It separates the three kinds of AI you will run into, shows where each one fits, and explains what changes when AI stops assisting operators and starts doing the routine work alongside them, as an AI teammate.
How is AI used in freight forwarding today?
Direct answer: AI in freight forwarding is used in three ways today: analytics that predict transit times, delays and rates; AI tools that speed up single steps such as document extraction or email drafting; and AI teammates that read operational emails, complete the routine work in your existing systems and escalate exceptions to an operator.
When people talk about AI in freight forwarding, or "AI freight forwarding" tools, they usually mean one of these three quite different things. It helps to separate them, because they solve different problems and they are bought by different people.
| Type of AI | What it produces | Who still does the work | Typical examples |
| Analytics and prediction | A forecast or a recommendation | The planner or manager acts on it | Route optimization, ETA prediction, rate forecasting, risk scoring |
| AI tools and copilots | A faster step inside a task | The operator, who checks, copies and sends | Document extraction, email drafting, chat assistants, summaries |
| AI teammate | Completed work, plus a short list of exceptions | The AI handles routine work; the operator handles judgment calls | Pre-alert to shipment in the FMS, milestone monitoring, arrival notices, invoice linking |
1. Analytics and prediction
This is the AI most industry reports focus on. Models look across large volumes of shipment, carrier and market data to predict transit times, flag likely delays, suggest routes or estimate rates. The output is a number or a recommendation that a planner acts on.
It is valuable when you have the data and the scale to use it. It does very little for the operator on a ten-person import desk who is working through a full inbox before lunch.
2. AI tools and copilots
The second group speeds up individual steps. An extraction tool pulls fields from a bill of lading. A writing assistant drafts a reply. A chat assistant summarizes a long thread.
Each of these saves a few minutes. But the operator still opens the email, decides which shipment it belongs to, checks the extracted data, copies it into the freight management system, updates tracking and sends the notice. The task list stays the same length; each item just gets slightly faster.
3. AI teammates
An AI teammate takes an operational input and finishes the routine work behind it. When a pre-alert arrives, it identifies the shipment, checks the documents against each other, opens or updates the file in your FMS, starts monitoring the container and prepares the next customer communication. When something is missing, conflicting or high risk, it stops and hands the issue to the right operator with the source email, the documents and a recommended next step attached.
The difference is the output. A tool gives an operator a faster step. A teammate gives the operator completed work and a short list of things that genuinely need a person.
Why the inbox is where AI in freight forwarding pays off
A single ocean import shipment can generate a long chain of operational emails between pre-alert and delivery. Each one follows a similar pattern: read it, work out which shipment it belongs to, check it against what you already know, key the new information into the system, decide what happens next and tell the right people.
Most of those steps follow rules the team already uses every day. They are repeated across hundreds of shipments a month, and they are exactly the kind of work that fills an operator's day without using their experience.
The inbox is also where expensive mistakes start. A changed ETA buried in a carrier update, or a last free day nobody noticed, turns into demurrage and detention charges that can run into thousands of dollars on a single container. Analytics tools can predict that a vessel is late. They do not read the carrier email, update the shipment and warn the customer.
That is why AI in freight forwarding delivers the most day-to-day value when it works where the work arrives. Large logistics companies are already moving in this direction. McKinsey's May 2026 analysis of AI in freight logistics describes one company using dozens of AI agents to handle high volumes of check calls, order acceptances and invoice payments, the same kind of repetitive coordination that fills a forwarder's inbox.
What does an AI teammate actually do?
It helps to describe an AI teammate by the jobs it does, in the order it does them.
- Read. It reads operational emails, attachments and carrier updates as they arrive, including pre-alerts, arrival notices, pickup orders and invoices.
- Understand. It identifies the shipment, the event and the information that matter, and checks references across documents before anything is written anywhere.
- Act. It applies your freight workflows and operating rules to decide the routine next step, then completes it: creating or updating the shipment, writing verified data back to your existing FMS, and preparing or sending notices.
- Monitor. It keeps watch on milestones such as ETA changes, holds, last free day and empty return, so nobody has to check carrier portals by hand.
- Escalate. When information is missing, conflicting or carries real cost or risk, it brings the issue to the right operator with the full context and a recommended action.
- Continue. Once the operator decides, it picks the work back up and carries the shipment forward to the next milestone.
The last two steps are what make it a teammate rather than a black box. Good AI knows when to step back, and it makes the hand-off easy instead of leaving the operator to reconstruct what happened.
A morning on an import desk, with and without an AI teammate
Without an AI teammate, an operator starts the day with a full inbox. The first hour goes to triage: which emails are pre-alerts, which are carrier updates, which are customers asking where their cargo is. Each pre-alert means opening the attachments, checking container numbers against the house bill, keying the shipment into the FMS and setting up tracking. Arrival notices have to be rewritten into customer notices. Somewhere in the middle, a customer calls about a container whose last free day is tomorrow.
With an AI teammate, much of that is already done before the operator logs in. Overnight pre-alerts have been matched and the shipments opened in the FMS. Tracking has started. Routine arrival notices are prepared, and the ones your rules allow have gone out. The operator's queue shows a handful of exceptions instead of a wall of email:
- a container number on the pre-alert that does not match the house bill of lading,
- an arrival notice showing a charge that was not on the original quote,
- a customer asking to change the delivery address after pickup has been booked.
Each exception comes with the emails, documents and shipment history attached. The operator spends the morning deciding, calling the customer and fixing the real problems, instead of copying data between windows.
For how to decide which actions run automatically and which wait for review, see our guide to human-in-the-loop AI for freight forwarding.
Will AI replace freight forwarders?
This is the question behind a lot of the interest in AI in freight forwarding, and it deserves a direct answer: no.
The parts of forwarding that customers pay for are judgment and relationships. Resolving conflicting instructions, negotiating with a carrier when a vessel rolls, making a compliance call, keeping a difficult customer informed: these depend on experience and trust, and they remain people's work.
What AI takes over is the repetitive coordination around those decisions. That changes the shape of the job rather than removing it. Operators handle more shipments without the workload doubling. Growth depends less on hiring ahead of volume. And the time that used to go to re-keying goes back to exceptions and customers.
What does it take to get value from AI in freight forwarding?
Choosing AI for freight forwarders is only partly a technology decision. The forwarders who get results tend to have a few things in place:
- Access to where the work arrives. An AI teammate needs to read the operational inbox your team already uses, not a new mailbox people have to remember to forward to.
- A connection to your system of record. Value comes from writing verified data back into your existing FMS, not from producing another spreadsheet.
- Your rules, written down. SOPs for when a notice goes out, who approves what and which customers get special handling are what turn AI from generic to useful.
- Clear escalation owners. Every exception needs a person to land with.
- A narrow start. One team, one inbox and one workflow, such as ocean import pre-alerts, over a two-to-four-week pilot. Expand once the results are consistent.
If you are comparing vendors, our overview of what AI software for freight forwarders can actually automate covers the capability questions to ask.
How to think about the ROI of AI in freight forwarding
The return from AI in freight forwarding is easiest to see when you measure it on your own shipments rather than on industry averages. A pilot should track:
- Time per shipment spent on email-driven tasks such as intake, updates and notices, before and after.
- Shipments per operator, which shows whether the team can absorb more volume without new hires.
- Avoided charges, such as last free days and detention windows caught in time.
- Corrections and overrides, which show how often the AI's work had to be fixed and whether that number falls over time.
- Response time to customers, which is often where customers notice the difference first.
Baseline these in the first week of a pilot. By the end of it you will have a return figure based on your own operation, not a vendor's slide.
Where NavLogic fits
NavLogic is the AI teammate built for freight operations. It connects to the inbox your team already uses (Gmail, Outlook or IMAP) and to your existing FMS, and it starts with ocean import:
- It reads pre-alerts and creates or updates the shipment in your FMS.
- It monitors ETA, holds, last free day and empty return across carrier updates.
- It prepares arrival notices and pickup instructions, and sends them at the right milestone.
- It reads carrier and D&D invoices, categorizes the charges and links them to the right shipment.
- It links every customer email to the shipment it belongs to.
Each workflow can run with human review, automatic sending or API hand-off, so you decide where the AI acts on its own and where an operator signs off. Exceptions come to the right person with the context attached.
NavLogic was founded by Grandy Li after more than 20 years building a freight forwarding company from a single office into a global operation. It is the software he wanted his own team to have.
Frequently asked questions
What is AI in freight forwarding?
AI in freight forwarding is the use of machine learning and language models to predict, assist with or complete forwarding work. It ranges from analytics that forecast transit times and rates, to tools that extract data from documents, to AI teammates that read operational emails, update shipments in the FMS, monitor milestones and escalate exceptions.
What is the difference between an AI tool and an AI teammate?
An AI tool speeds up one step of a task that an operator still owns, such as extracting fields from a document. An AI teammate owns the routine task from start to finish: it reads the input, decides the next step under your rules, completes the work in your systems and hands exceptions to a person with context.
Will AI replace freight forwarders?
No. Judgment, carrier and customer relationships, and compliance decisions remain people's work. AI takes over repetitive coordination such as data entry, tracking checks and routine notices, which lets each operator handle more shipments and spend more time on exceptions and customers.
Is AI only useful for large freight forwarders?
No. Predictive analytics tends to favor large networks with a lot of data, but AI teammates work at the level of a single team and inbox. Small and midsize forwarders often feel the benefit sooner, because each operator covers more workflows and has less time to spare.
What is the ROI of AI in freight forwarding?
It depends on your volume and workflows, so it is best measured in a pilot. Track time per shipment on email-driven tasks, shipments per operator, avoided demurrage and detention charges, correction rates and customer response times before and after.
Is it safe to give an AI teammate access to an operations inbox?
It should be, if the system is built for it. Look for clear permissions on what the AI can read and change, a full audit trail of every action, review modes for customer-facing messages, and escalation whenever information is missing or conflicting. Ask any vendor how your data is stored and whether it is used to train shared models.
Bring us your hardest email
In 30 minutes, see how NavLogic fits your workflows, inbox and existing systems.
Start Free
