AI in freight forwarding, 2026: what is paying off and what is still a slide deck

Technology · October 15, 2026 · 5 min read

The gap between AI ambition and AI results in logistics is now measurable. Boston Consulting Group's 2026 survey of 30 leading logistics companies found 97% rank AI as a strategic priority and only 13% can point to measurable financial impact. Yet the largest forwarders are now putting AI savings into investor guidance, and the use cases that pay are consistent: quoting, reading the inbox, extracting documents and, newest, letting customers' own AI agents act through APIs. For a mid-sized forwarder in Latin America or the US, the lesson is to stop shopping for AI as a product and start fixing the data and workflows it runs on.

The scorecard: priority everywhere, payback rare

BCG's August 2026 analysis of its executive survey found 70% of logistics companies have an AI strategy and 67% a dedicated budget, but only 13% report measurable financial impact. Asked for their most impactful AI solution, 80% named a single standalone tool. The top barrier cited, by 37%, was lack of internal AI expertise, followed by fragmented legacy systems that cannot share data. And 93% see a notable risk from AI-native competitors.

A wider BCG and Alpega survey run in January 2026, covering 84 logistics providers and 98 shippers, found 40% of providers had deployed AI beyond pilots but only 10% had embedded it in core operations. Nearly half planned AI in pricing, quoting and customer service. More than 40% of shippers now consider a provider's AI capability when choosing partners, though fewer than 10% treat it as mandatory.

What is working: quoting and the inbox

The clearest evidence is in quoting. One of the largest global forwarders told investors in May 2026 that it now reads requests for quotation from email with AI using intent classification, feeds structured inputs into its internal quoting system, which handles about 1.6 million quotes a year, and uses AI agents to run follow-up cycles. It reports quote preparation down from about 30 minutes to seconds and a hit-rate lift of at least 10%.

Another global forwarder has gone further in public guidance. It expects a nine-figure annual profit gain from AI by the end of 2027, from a 5% productivity gain across a white-collar workforce it sizes at more than 25,000 addressable full-time equivalents. It cites air freight quotes delivered twice as fast, sea bookings cut from minutes to seconds and custom EDI or API connections completed in under a day. A third puts its AI target at about US$1.4 billion in productivity value by 2030.

  • Quote intake: classifying and extracting RFQs from email into structured fields.
  • Pricing: applying rate cards to those fields so a person reviews instead of builds.
  • Follow-up: agents chasing open quotes on a schedule.
  • Booking and data enrichment: filling shipment records from documents and messages.

Agents and MCP: from chatbot to API caller

The shift in 2026 is from AI that answers questions to AI that takes actions through systems. Gartner forecast in April that spending on supply chain software with agentic AI will grow from under US$2 billion in 2025 to US$53 billion in 2030, and that 60% of enterprises using such software will have adopted agentic features by 2030, up from 5% in 2025. One of the largest express carriers has said it plans AI agents in more than half of its operational workflows by 2028, organised as manager, worker and audit agents.

The Model Context Protocol is how many of these agents reach a forwarder's systems. In October a US forwarder announced an MCP connection that lets customers quote, track and book through compatible AI assistants, with the customer confirming through the forwarder before a booking is made and complex pricing routed to its staff. That human checkpoint is the pattern to copy: let outside agents read and request, and keep commitments behind your own approval.

Where the hype still runs ahead

Gartner warned in June 2025 that more than 40% of agentic AI projects will be cancelled by the end of 2027 because of cost, unclear value or weak risk controls, and estimated that only about 130 of the thousands of vendors claiming agentic AI deliver genuine agentic capability. It calls the relabelling of chatbots and robotic process automation as agents agent washing.

Inside forwarding the debate has moved to architecture. The Loadstar reported in August that the industry argument now pits a single broad super agent against clusters of specialist agents, with one view that the models are already capable enough and data quality is the real limit. That matches BCG's finding: isolated tools on fragmented data do not pay. An agent that cannot see your rate cards, job history and documents in one place will guess, and guessing at a freight quote is expensive.

The Latin American angle

None of the surveys above breaks out Latin American forwarders, and we found no reliable regional adoption figure, so be wary of anyone quoting one. What the evidence does show is that the gains come from email-heavy, repetitive work, which describes most LATAM and cross-border US operations: spot quotes in Spanish, Portuguese and English, PDFs from overseas agents, and customers who expect WhatsApp-speed answers.

The constraint is usually not the model but whether your rates, customers and jobs live in one system the AI can read. A forwarder with clean rate cards and structured job data can deploy quoting assistance in weeks. One running on spreadsheets and inboxes will spend that time on data cleanup first.

Sources

  1. Why AI Isn't Delivering ROI in Logistics in 2026Boston Consulting Group · August 24, 2026
  2. AI Is Already Moving the Logistics Industry ForwardBoston Consulting Group · March 27, 2026
  3. AI has reached forwarders' P&L, now the arguments beginThe Loadstar · August 12, 2026
  4. Agentic AI supply chain software to hit $53 billion by 2030MH&L (citing Gartner) · April 13, 2026
  5. Over 40% of Agentic AI Projects Will Be Scrapped by 2027, Says GartnerOutlook Business · June 25, 2025