WW/WAREHOUSIN

Filed 672W3M read

57% expect autonomous AI for over a quarter of warehouse decisions

A survey of nearly 2,000 logistics professionals across 21 countries finds broad AI adoption, limited quantified gains and more autonomous warehouse decisions.

By
Amara Osei
Filed
Length
672 words
Read
3 min
AI could make a quarter of warehouse decisions autonomously by 2028: study
AI could make a quarter of warehouse decisions autonomously by 2028: studycegoh / pixabay

Key points05

  • 57% expect AI to make or execute more than 25% of warehouse decisions autonomously by 2028; the figure reaches 70% among fully integrated organisations.
  • The study surveyed nearly 2,000 supply chain and warehousing professionals across 21 countries.
  • 61% report large-scale or full generative AI deployment, compared with 56% for agentic AI.
  • AI agents can act without human approval on inventory movements in 51% of agentic use cases, data corrections in 42% and financial decisions in 12%.
  • 90% expect higher AI investment over the next two to three years; 29% plan increases exceeding 25%.

By 2028, 57% of organisations expect artificial intelligence to make or execute more than 25% of operational warehouse decisions autonomously. The research comes from the Intelligent Logistics Systems Lab at MIT’s Center for Transportation & Logistics and warehouse automation specialist Mecalux, which surveyed nearly 2,000 supply chain and warehousing professionals across 21 countries.

Among organisations that have fully integrated the relevant technologies, the expectation rises to 70%. That is 13 percentage points above the overall result, a difference that links full integration with greater expected autonomy without proving causation.

Adoption already runs ahead of the autonomy forecast. Sixty-one percent report large-scale deployment or full integration of generative AI, compared with 56% for agentic AI.

Most applications remain new. Around 75% said their AI applications had stayed at their current stage for less than 12 months, making the 2028 forecast part roadmap and part conditional promise.

Evidence of value remains mixed

Among use cases respondents described in detail, 87% of generative AI applications and 81% of agentic AI applications produced at least moderate improvements in their target performance metrics. Only 39% and 36%, respectively, had achieved significant gains that organisations quantified and documented.

Agents have operational authority, not financial carte blanche

Respondents said AI agents could make operational decisions without human approval, most often for inventory movements, which account for 51% of agentic AI use cases. Data corrections account for 42% of those cases, while just 12% permit autonomous financial decisions.

Integration is the constraint

Forty percent of respondents named legacy-system integration as the biggest obstacle to scaling generative AI. A majority, 54%, said better integration with warehouse management and ERP systems would accelerate deployment.

What changes for shippers, carriers and forwarders?

The immediate change is control, not transport capacity. The study measures decision authority, application performance and systems integration; it makes no claim about freight rates, transport capacity or transit times.

For warehouse operators and shippers, delegated inventory movements and data corrections could speed routine processing and improve record quality. Carriers and forwarders may eventually receive fewer manual data-correction handoffs, but that remains an inference, not a documented service or cost gain.

The evidence supports controlled expansion rather than blanket autonomy. Nearly four in ten generative AI cases and just over one in three agentic cases had documented significant gains, while financial authority remained uncommon. The technology is earning responsibility in increments, not by declaring independence.

Matthias Winkenbach, director of MIT’s Intelligent Logistics Systems Lab, said: “The objective is to move beyond general sentiment and to provide practical guidance on where these technologies are delivering operational value today and where they remain experimental.” That distinction is where budgets get interesting.

Investment is set to rise

Ninety percent of organisations expect to increase AI investment over the next two to three years. Within that group, 29% plan increases of more than 25%, indicating substantial budget growth across a large part of the sample.

That does not guarantee returns. The combination of high deployment, short application histories and limited quantified gains suggests buyers are funding capability before the performance ledger is complete. Software, apparently, still wants footnotes.

Warehouse work is changing

Fifty-six percent of organisations said AI had “mainly shifted employees towards higher-value tasks.” Only 11% reported reductions in headcount, so role changes outnumbered reported cuts by more than five to one.

The reskilling effort is broader still: 73% of organisations are upskilling warehouse-floor staff, and 61% are creating new technical roles. AI is changing jobs before it is removing them, at least in this sample.

AI in warehousing will be a key theme at IntraLogisteX on 17-18 March 2027 at the NEC Birmingham. Mecalux will exhibit, and the full conference programme is available through the event’s official website.

The next two to three years will test whether rising AI investment produces documented operating gains, not merely more pilots and permissions. No freight-rate or capacity shift can be read from this study; for buyers, documented performance is the practical signal worth watching.

Original: intralogistex.co.uk

Share this article:

More from Amara Osei

Amara Osei

Show full bio

Staff writer covering regulation, policy and their impact on Ocean and air freight, container shipping at Waybill Wire.

961 articles

Related05

  1. 57% of Firms Expect AI to Run a Quarter of Warehouse Decisions by 2028

  2. 90% of warehouse operators plan more AI spend, but ROI proof lags

  3. AI to handle more than a quarter of warehouse decisions by 2028

  4. 83% of supply chain leaders say their workforce can't use AI

  5. Just 18% see better customer experience from port AI, BCG study finds

« Prev