Nearly half of warehouses have adopted AI to improve visibility and reduce manual work, but only 5% trust it enough to make decisions without human review. Most organizations use AI to inform and recommend actions while employees retain final decision-making authority.
- 49% of warehouses are running AI in at least one process, piloting in multiple areas, or have embedded it across operations
- Only 5% of respondents trust AI enough to act on recommendations without review, while 83% usually review AI recommendations first
- AI delivers greatest value through operational visibility (33%), reduced manual work, and early issue identification
- Top obstacles include budget constraints (25%) and data quality issues (22%)
- 30% of leaders view autonomous decision-making as the greatest untapped potential for warehouse AI
AI has moved well past the pilot stage in many warehouses, but it is still mostly informing people rather than acting on its own, according to new data from Logistics Reply, a Reply Group company.
In fact, nearly half of respondents (49%) said AI is already running in at least one warehouse process, piloting in a couple of areas, or embedded across much of their operation.
Another 28% are actively evaluating where it could add value, while 23% have not started engaging with AI yet. Very few organizations describe themselves as AI skeptics; most have either adopted the technology or are actively working toward it.
"Warehouse operators have moved beyond the question of whether AI belongs in the warehouse and into a more important one: how AI can create greater measurable operational value," says Michelle Jones, director of presales and solutions consulting. "At this stage, AI is delivering the greatest value by improving operational visibility, supporting better decisions, and reducing manual work. The opportunity ahead is to move from AI that informs teams to AI that can recommend, coordinate, and ultimately orchestrate action across the warehouse."
Key takeaways:
· Reporting and analytics (32%) and slotting or replenishment (27%) were the two most common use cases, followed by order fulfillment and yard or dock management.
· Asked what AI actually does in their warehouse right now, 31% of respondents said it generates reports, dashboards, or operational insights, while only 3% said AI makes operational decisions on its own.
· When an operational disruption hits, such as an inventory shortage, a labor gap, or an equipment failure, roughly 48% of respondents said AI flags the issue while people coordinate the response manually, and a nearly identical share said AI recommends the next best action while people make the final call. Just 4% said AI automatically coordinates the response across systems on its own
· Most respondents (83%) said they usually review AI recommendations before acting on them, and only 5% said they trust AI enough to act without review. When AI does make a recommendation or decision, 75% of respondents said employees can override it, either with approval or depending on the process.
· More than half of respondents (58%) said their AI coordinates work across a few connected operational systems, while 35% said it primarily operates within a single application. Only 1% said AI currently orchestrates workflows across multiple systems, such as WMS, ERP, TMS, labor, and automation, together.
· Autonomous decision making was named the area with the greatest untapped potential (30%), ahead of inventory optimization and labor planning. Looking forward, the top capability leaders want from AI is for it to recommend the best operational actions (23%), closely followed by orchestrating decisions across multiple systems (22%).
· Respondents pointed to better operational visibility (33%) as the single biggest business value AI has delivered so far, followed by reduced manual or repetitive work and earlier identification of issues or disruptions. But progress is not without friction: budget constraints (25%) and data quality (22%) were named the top obstacles to getting more value from AI, ahead of difficulty proving ROI and a lack of internal expertise.
