Study finds AI will run more warehouse decisions by 2028
Fri, 20th Nov 2026 (Today)
A study by the MIT Centre for Transportation & Logistics and Mecalux found that 57% of organisations expect AI to make or execute more than a quarter of warehouse operational decisions autonomously by 2028. The survey covered nearly 2,000 supply chain and warehousing professionals across 21 countries.
The report examined the use of generative and agentic AI in warehouse operations and found adoption is already widespread. It found that 61% of organisations reported using generative AI at large-scale deployment or full integration, compared with 56% for agentic AI.
Most applications remain relatively new, with about three in four deployments at their current stage for less than a year.
Even so, respondents reported early gains from the technology. Among the use cases described in detail, 87% of generative AI applications and 81% of agentic AI applications showed at least moderate improvement in their target metrics. However, only 39% of generative AI use cases and 36% of agentic AI use cases had significant gains that were both quantified and documented.
Where respondents were able to estimate the scale of improvement, the typical gain was about 20% on the main target metric.
Autonomy limits
The research suggested companies are giving AI systems decision-making authority in selected parts of warehouse operations while keeping tighter controls elsewhere. Inventory movements were the area where AI agents were most often allowed to make decisions without human approval, accounting for 51% of agentic AI use cases, followed by data corrections at 42%.
By contrast, only 22% of these systems were allowed to make customer-facing commitments, and 12% were permitted to make financial decisions. The most common oversight model was "human-on-the-loop", in which AI operates within defined limits while employees monitor activity and can intervene.
"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," said Matthias Winkenbach, Director of the Intelligent Logistics Systems Lab.
Governance effect
The report also pointed to a link between governance practices and stronger reported outcomes. Organisations with comprehensive, actively enforced AI governance were about twice as likely to report significant, documented improvements.
For generative AI, 51% of companies with such governance reported significant and documented gains, compared with 26% of those without it. For agentic AI, the figures were 46% and 23%.
This pattern remained after accounting for deployment status, region, industry and budget, although the study noted that the relationship showed association rather than proof of causation.
Integration with existing systems emerged as the main barrier to broader use. Forty per cent of respondents cited legacy system integration as the biggest constraint on scaling generative AI, while 54% said better integration with warehouse management and ERP systems would have the greatest effect on adoption.
Investment plans
Most organisations said they expect spending on AI to rise over the next two to three years. The study found that 90% anticipated an increase in investment, with 29% expecting that rise to exceed 25%.
That was broadly in line with the previous year's study, which found that 87% expected higher spending. Among companies that had fully integrated the technologies, 70% said they expected agentic AI to make or execute more than a quarter of warehouse operational decisions autonomously by 2028, above the overall figure of 57%.
The country findings highlighted stronger large-scale adoption of agentic AI in the Netherlands, Belgium, Australia, France, Canada, the United Kingdom and Germany.
Workforce shifts
On employment, the research suggested AI is changing roles more often than reducing headcount. Fifty-six per cent of organisations said the technology had mainly shifted staff towards higher-value tasks, while 11% reported reduced headcount and 9% reported increased headcount.
Companies also reported investment in workforce changes alongside technology roll-outs. Seventy-three per cent said they were upskilling warehouse-floor staff, and 61% said they were creating new technical roles.
The survey was based on 1,967 valid responses collected online from C-suite executives and senior managers at organisations with at least 100 employees and their own warehouse operations.
"Warehouse decision-makers need AI they can measure. This research gives our clients a clear map of where the technology is working today and what it takes to prove it," said Javier Carrillo, Chief Executive Officer, Mecalux.