--- title: "Why employees don't use AI tools, and how to find out | Spradley" description: "Employees skip AI tools for practical reasons: no use case, no time, unclear rules, a poor first answer or fear. How to find out which apply in your teams." url: https://spradley.io/articles/why-employees-dont-use-ai-tools --- # Why employees don't use AI tools, and how to find out Employees skip AI tools for practical reasons: no use case, no time, unclear rules, a poor first answer or fear. How to find out which apply in your teams. Miguel Cabral Oliveira Co-founder & CTO Aug 6, 2026 ![A man writes in a paper ledger beside an unopened cardboard box on his desk](https://spradley.io/media/blog/why-employees-dont-use-ai-tools/hero.webp) Employees usually stop using AI tools for practical reasons. The tool has no clear use in their job, they have no time to learn it, the rules are unclear, the first answers were poor, or they worry that using it looks lazy or puts their job at risk. Which of these applies differs from team to team, so ask people directly, with follow-up questions, and compare the answers by department. ## Why aren’t employees using the AI tools we rolled out? When usage stays low, leaders often read it as attitude: people don’t like change. The reason is usually more ordinary, and easier to fix. Six of them account for most of it. ### There is no obvious use for it in their job A licence and a launch email don’t tell a transport planner what to do with a chat assistant on Monday morning. The tools were often demonstrated on office tasks such as summarising meetings or drafting emails, which suit some roles far better than others. [BCG’s AI at Work 2025](https://www.bcg.com/publications/2025/ai-at-work-momentum-builds-but-gaps-remain) survey found regular use among frontline employees had stalled at 51%, well behind leaders and managers. If people can’t name one task the tool helps with, they won’t form a habit. ### They have no time to learn it Learning a new tool costs a few slow weeks before it saves anything. The [2025 Work Trend Index](https://www.microsoft.com/en-us/worklab/work-trend-index/2025-the-year-the-frontier-firm-is-born) reports that 80% of the global workforce, leaders and employees alike, say they lack enough time or energy to do their work. BCG found regular use was much higher among employees who had at least five hours of training, yet only a third say they have been properly trained. ### The rules are unclear If people don’t know whether they may paste a customer email or a patient note into the tool, many will simply not use it. [Gallup’s 2025 research](https://www.gallup.com/workplace/691643/workplace-trends-watch-2025.aspx) found that 44% of employees said their organisation had begun integrating AI, while only 22% said it had communicated a clear plan, and 30% said there were general guidelines or formal policies for using AI at work. ### The first answer was poor Someone in payroll asks about overtime rules in the local agreement and gets a confident, wrong answer. They tell two colleagues at lunch. Nobody in that team tries again. A bad first week does more damage than any training session can repair, because people stop giving the tool a second chance. ### They worry it looks lazy, or makes them replaceable Some people use the tools and keep quiet about it. In the [2024 Work Trend Index](https://www.microsoft.com/en-us/worklab/work-trend-index/ai-at-work-is-here-now-comes-the-hard-part), 52% of people who use AI at work said they were reluctant to admit using it for their most important tasks, and 53% worried it makes them look replaceable. Low usage figures can partly reflect people using personal accounts out of sight, which we cover in our post on [shadow AI](https://spradley.io/articles/shadow-ai). ### Their manager doesn’t use it People follow what their manager does more than what the launch email says. Gallup’s [AI indicator](https://www.gallup.com/699797/indicator-artificial-intelligence.aspx) calls manager support the strongest driver of widespread adoption: employees who strongly agree their manager actively supports AI use are 1.7 times as likely to use it a few times a week or more. BCG reports that the share of employees who feel positive about generative AI rises from 15% to 55% with strong leadership support, and that only about a quarter of frontline employees say they get it. ## Why do employees resist AI at work? Most of what gets called resistance is one of the reasons above. Fear of job loss is real, though it isn’t only a frontline worry. BCG found leaders and managers were more likely than frontline employees to worry about losing their job in the next ten years (43% against 36%), and that people in companies redesigning work around AI worried more (46%) than those in less advanced companies (34%). That matters for how you respond. If a team holds back because they have no use case, a workshop built on their own tasks helps. If they hold back because they fear what the tool means for their role, more training won’t move them. They need a straight answer from leadership about what the tools are for. ## How do we find out why, in our own organisation? You can’t tell these reasons apart from usage data. A team with low activity might have no use case, an unclear policy or a manager who never mentions it. You have to ask. 1. Use your usage reports to find where adoption is low or dropping. Our guide to [measuring Copilot adoption beyond licence data](https://spradley.io/articles/measure-copilot-adoption-beyond-licence-data) explains what those reports can show. 2. Ask people in those teams, and in a few teams where use is high, about their last real attempt to use the tool. Open questions with a follow-up work better than a rating; there is a copyable list in our post on [AI adoption survey questions](https://spradley.io/articles/ai-adoption-survey-questions). 3. Make it safe to be honest. Say who will see the answers and report only for groups big enough that no one can be picked out. People won’t admit they use a personal account, or that they fear for their job, if they think their manager will read it with their name on it. 4. Compare departments. The reasons in a warehouse office won’t match those in the legal team, and the company average hides both. 5. Tell people what you heard and what will change, then ask again a quarter later. Spradley does steps two to four for you. It runs one-on-one AI interviews by text or voice, sent by email or Slack, that people answer on their own phone or computer in their own time. The interviewer follows up on each answer, so “I don’t really use it” turns into the reason. The report groups answers into themes with the quotes behind them, lets you compare departments, and never shows numbers or ratings for fewer than five people. More on our [AI adoption page](https://spradley.io/solutions/ai-adoption). ## What can leadership do about it? Once you know the reasons, the fixes are fairly direct. - Name two or three tasks per role where the tool helps, and show them on that team’s own work. - Give protected time to learn, and count it as work. - Publish a short, plain policy that says what data can go in and what can’t. - Ask managers to use the tools openly and talk about what didn’t work as well as what did. - Say clearly what the tools are for, including what they mean for roles, so people aren’t left to guess. Then check whether it worked by asking the same people again. ## Frequently asked questions ### Why are employees not using AI tools? The common reasons are practical: no clear use in their job, no time to learn, unclear rules on what is allowed, a poor first experience, or worry that it looks lazy or threatens their role. Which ones apply varies by team. ### Can usage data tell us why adoption is low? No. Usage reports show who used the tool and how often. They can point you to the teams where use is low, but you need to ask people to learn the reason. ### Is low AI usage a sign of resistance to change? Usually not. What looks like resistance is more often a missing use case, unclear policy or lack of time. Fear for jobs is real, and it needs a direct answer from leadership rather than more training. ### How do we get honest answers about AI use? Ask open questions about real tasks, say who will see the answers, and only report for groups large enough that no one can be identified. ## Key terms use case A specific task in someone’s job where a tool helps. Without one, people rarely form a habit of using it. shadow AI Employees using AI tools for work that their organisation has not approved, such as a personal chatbot account, a browser extension or an app on their own phone. minimum group size The smallest number of responses a group needs before its results are shown. Five is a common default; smaller groups are hidden or combined with a larger one. AI-moderated interview A one-on-one conversation run by software, in which the interviewer asks open questions and a follow-up question based on each answer. Everyone invited can take part at the same time. theme with quotes A pattern found across many answers, reported together with the quotes it came from so readers can check how it was drawn. ## More articles [Explore all resources](https://spradley.io/articles) [ ![A man with a clipboard and tally counter watches colleagues walk through an office](https://spradley.io/media/blog/measure-copilot-adoption-beyond-licence-data/hero.webp) AI adoption ### How to measure Copilot adoption beyond licence data What the Copilot usage report and Copilot Dashboard show, what they miss, and how to measure AI adoption by use, proficiency and the reasons behind both. ](https://spradley.io/articles/measure-copilot-adoption-beyond-licence-data)[ ![Two office workers pin rows of index cards to a cork wall, one standing on a chair](https://spradley.io/media/blog/ai-adoption-survey-questions/hero.webp) AI adoption ### AI adoption survey questions to ask employees 24 open AI adoption survey questions to ask employees about your rollout, grouped by purpose, and why a follow-up question beats a 1 to 5 scale. ](https://spradley.io/articles/ai-adoption-survey-questions)[ ![An office worker types on his own typewriter under the desk while the office computer sits covered](https://spradley.io/media/blog/shadow-ai/hero-v2.webp) AI adoption ### Shadow AI: why employees use unapproved AI tools Shadow AI is staff using AI tools you haven't approved, like personal ChatGPT accounts. Why it happens, why bans fail, and how to find out without blame. ](https://spradley.io/articles/shadow-ai)[ ![A manager in an insurance office, seen from behind, taps a blank wall barometer while a claims handler at a desk of folders raises a hand to speak](https://spradley.io/media/articles/spradley-vs-lattice/hero.webp) Comparisons ### Spradley vs Lattice for Employee Feedback Spradley vs Lattice: run one fallen score through both, see what each does when the reason is unknown, and when to keep Lattice and add interviews. ](https://spradley.io/articles/spradley-vs-lattice)[ ![A practice manager in a 1980s veterinary clinic, seen from behind, presses a stethoscope to the closed staff-room door while a vet nurse in profile waits beside her holding a cat carrier](https://spradley.io/media/articles/spradley-vs-leapsome/hero.webp) Comparisons ### Spradley vs Leapsome: Surveys or Reasons? Spradley vs Leapsome on one workplace problem: what each tells you, what each does when the cause is still unknown, and whether to keep, add or replace. ](https://spradley.io/articles/spradley-vs-leapsome)[ ![A 1980s airport baggage hall. A manager in a blazer, seen from behind, kneels with a torch to peer under a stopped luggage carousel, while a baggage handler in ear defenders, seen in profile, stands holding a suitcase in each hand, waiting to load them](https://spradley.io/media/articles/what-to-do-when-one-department-has-low-engagement-scores/hero.webp) Taking action ### One Department Has Low Engagement Scores: What to Do Is it the manager, the work, or a company problem that only shows there? A decision rule, the evidence for each, and what to do when the group is tiny. ](https://spradley.io/articles/what-to-do-when-one-department-has-low-engagement-scores)