What ecommerce businesses trust AI to do, and where they draw the line
Ecommerce businesses trust AI most with work they can check. In the Linnworks Market Survey 2026, 57% of 500 US and UK ecommerce decision-makers said they would trust AI to generate demand forecasts on their behalf. Only 36% would trust it to make purchasing decisions, and 36% to communicate with suppliers.
That gap tells you a lot about how to bring AI into your operation. This guide covers what businesses trust AI with, why the line falls where it does, and how to expand what AI handles without risking orders or spending.
What do ecommerce businesses trust AI to do?
Demand forecasting is the task ecommerce businesses trust AI with most, at 57%. Drafting product listings and flagging inventory or fulfillment problems follow at 50% each. Trust falls to 36% for the two tasks that commit the business: making purchasing decisions and communicating with suppliers.
- Generate demand forecasts 57%
- Draft or optimize product listings 50%
- Flag inventory or fulfillment problems 50%
- Answer questions about business data 45%
- Reroute or prioritize orders automatically 43%
- Make purchasing or reordering decisions 36%
- Communicate with suppliers 36%
Almost nobody rules AI out entirely. Only about 1% say they wouldn’t trust AI to do anything on their behalf.
Why the line falls where it does
Trust follows how easy it is to check or undo a result. A demand forecast is a recommendation. If it’s wrong, a planner can spot it before anyone places an order. A purchase order is a commitment. If it’s wrong, the money is spent, the stock is on its way, and a supplier is involved.
The same logic explains where AI already acts on its own. 59% of decision-makers say they have an AI assistant taking actions on their behalf, such as sending messages, updating records, or moving data between systems. Those are low-stakes actions that are easy to reverse.
This caution sits alongside wide adoption. 52% of ecommerce businesses describe AI as scaled or broadly adopted across their organization, 45% are piloting or exploring it, and only 3% have chosen not to use it. Businesses are using AI. They are choosing carefully where to let it act.
Data is part of that choice. 28% say they have become more cautious about AI on data and security grounds, rising to 34% in the US. And integration complexity is the most common barrier to adopting more AI in operations (18%), ahead of an unclear return on investment (8%). AI that only sees part of the picture makes recommendations a team has to double-check, which is exactly the kind of task businesses already keep for people.
How to widen what AI handles in your operation
Start where trust is already high, then widen what AI handles as it proves itself.
- Connect your data first. Forecasts and alerts are only as good as the orders, stock, and channel data behind them. Get inventory and orders into one system before you judge any AI recommendation.
- Start with recommendations a person reviews. Demand forecasts, reorder suggestions, alerts about stock or fulfillment problems, and draft listings all let your team check the output before anything happens.
- Measure accuracy for a set period. Track how often your team overrides a recommendation, and why. Three months of those records will show which recommendations you can rely on and which still need checking.
- Let AI act on reversible tasks. Tagging orders, sorting them into folders, or routing them using rules your team sets are all easy to check and easy to undo. Just keep a log of what rules ran.
- Set guardrails before delegating spending. Before AI sends purchase orders, agree on spend limits, approved suppliers, and the order sizes that still need a person to sign off.
How Linnworks keeps your team in control of AI
Linnworks is built around the same order of trust. Orders, inventory, listings, and shipping run through one system across 100+ sales channels, so AI recommendations draw on complete data.
Spotlight AI watches for the manual actions your team repeats, such as tagging orders or assigning shipping services, and suggests automation rules to replace them. Your team reviews each suggestion before anything runs. Users say they save 30+ hours a month on top of the savings they get from rules they set up manually.
Stock forecasting predicts demand by SKU and surfaces reorder recommendations, so purchasing decisions stay with your team, backed by better numbers.
For more on when AI starts to pay off, see why AI becomes essential once ecommerce hits scale.
See how your operation compares
Where does your business sit on the trust curve? The Linnworks Market Survey 2026 benchmark asks five quick questions about your order mix, automation, AI, and EDI, then shows how you compare with 500 ecommerce decision-makers, with a practical next step for each result.
FAQ
Demand forecasting is the task ecommerce businesses trust AI with most, at 57%, according to the Linnworks Market Survey 2026. Half would trust AI to draft or optimize product listings and to flag inventory or fulfillment problems.
Only 36% of ecommerce decision-makers would trust AI to make purchasing or reordering decisions on their behalf today, and the same share would trust it to communicate with suppliers. These are the tasks businesses trust AI with least.
52% of ecommerce businesses describe AI as scaled or broadly adopted across their organization, and 45% are piloting or exploring it. Only 3% have chosen not to use AI.
Start with recommendations your team can review before they take effect, such as demand forecasts and stock alerts. Connect your order and inventory data first, measure how accurate the recommendations are, and set spending limits before letting AI make purchasing decisions.
Source: Linnworks Market Survey 2026, conducted by OnePoll among 500 US and UK ecommerce decision-makers, July 28 – August 4, 2026.