# How to Get Skeptical Employees to Use AI in a Small Business Canonical URL: https://business.theaieducator.io/posts/how-to-get-employees-to-use-ai Publication: Dan Fitzpatrick Insights Author: Dan Fitzpatrick Topic: AI Adoption and Change Published: 2026-09-27T15:44:36.000Z Modified: 2026-09-27T15:44:39.108Z Most staff who avoid AI are not afraid of it; they prefer their own way of working. Start with the chore they would happily give away, never the craft they are proud of, and let the hours saved persuade the rest. ## In brief To get skeptical employees to use AI, start with the chore they would happily give away, never the craft they are proud of. Ask each reluctant person which part of their week they would drop, put AI there first, let them overrule anything it produces, and measure hours handed back, not logins. ## Key takeaways - The most common reason employees give for not using AI is a preference to keep working the way they already do: 46% of non-users in Gallup's February 2026 survey of 23,717 US workers. - Manager support, clear guidelines and a good fit with existing work all go with more frequent AI use in Gallup's data; pep talks do not appear on the list. - Some doubt is hidden: 30% of people at US businesses with 2 to 250 employees said they sometimes act more optimistic about AI among colleagues than they really feel. - Pressure can push resistance out of sight. Harvard Business School researchers describe 'symbolic adoption', where employees appear to use AI without really doing so. - The Chore Rule: start with the chore a reluctant person would happily give away, never the craft they are proud of, leave the judgment and final say with them, and let what they save persuade the next person. - The most common mistake is leading with the craft: demonstrating AI doing someone's best work reads as a threat, not an invitation. - Measure hours handed back on a named job, reported by the person doing it, rather than counting who logs in. You get skeptical employees to use AI by starting with the chore they would happily give away, never the craft they are proud of. Ask each reluctant person which part of their week they would drop tomorrow, put AI on that job first, leave every judgment and the final say with them, and let the hours they save persuade the next person. Mandates, league tables of who logs in and demonstrations of AI doing someone's best work all tend to make resistance quieter, not smaller. That last point matters more than it sounds. In most small businesses the person who has not touched the new AI tool is not a technophobe. It is often the most experienced person you employ, the one whose way of working is the reason customers come back. Getting them on board is not a training problem. It is a question of what you ask AI to take away from them first. ## Why do employees refuse to use AI? The most common reason is not fear or ignorance: people who do not use AI mostly prefer to keep doing their work the way they already do it. Gallup surveyed 23,717 employed US adults in February 2026 and asked the people who do not use AI at work why not. The top answer, given by 46% of non-users, was a preference to keep doing work the way it is currently done. Close behind came ethical opposition to AI (43%), concerns about data privacy and security (43%) and a belief that AI cannot help with their work (39%). Even among people who use AI only occasionally, 36% gave the same "I prefer my way" answer ([Gallup, April 2026](https://www.gallup.com/workplace/704252/workplace-separates-adopters-holdouts.aspx)). Read that list as an owner and something useful appears. Only one of the four reasons is about AI itself. The biggest is about the person's work: they have a way of doing it that works, that they built, and that nobody has given them a reason to change. Gallup's data also shows what moves people. Employees who strongly agree that their manager actively supports AI use were frequent users 78% of the time, against 44% for everyone else. Where there were clear guidelines for AI use, 68% were frequent users, against 47% where there were not. And where AI fitted the systems and processes people already used, 88% were frequent users, against 55% where it did not. Support, rules and fit with the real work: none of those is a pep talk. ## Is your team really on board, or only saying so? Some of your team are probably more doubtful about AI than they let you see. In a survey of 1,009 people working in US businesses with 2 to 250 employees, published in January 2026, 30% agreed with the statement "Sometimes, I act more optimistic about AI among colleagues than I really feel." Almost as many, 31%, said they sometimes feel their job is less meaningful when machines handle challenging work ([Business.com, January 2026](https://www.business.com/articles/ai-usage-smb-workplace-study/)). Harvard Business School researchers Das Narayandas and Shunyuan Zhang have a name for what happens next. Employees who feel AI threatens the part of the job that gives them standing rarely refuse outright. As the Harvard Business School write-up of their work puts it, "People don't want to get adversarial or be identified as the curmudgeon who refuses to do things, so they adopt the technology symbolically, giving the impression they are using it" ([HBS Working Knowledge, June 2026](https://www.library.hbs.edu/working-knowledge/why-employees-resist-ai-and-how-companies-can-win-them-over)). They name the threats too. One is role compression, where judgment and expertise are automated along with the daily tasks, and the duties left over feel lower in status. Another is control shift, where decisions that once defined someone's expertise move to an algorithm. Push harder and the resistance can move out of sight. In a survey of 2,400 knowledge workers and executives at companies with 100 to more than 10,000 employees, carried out for the AI company WRITER, 29% of employees admitted to sabotaging their company's AI strategy, for example by entering company information into public tools, using unapproved tools or refusing to use AI. In the same survey, 60% of executives said they plan to lay off employees who can't or won't use AI ([WRITER, April 2026](https://writer.com/blog/enterprise-ai-adoption-survey-results-press-release/)). Those are larger companies than most readers here run, and the survey does not show that one figure causes the other. My reading is simpler: when using AI feels like a test of loyalty, people pass the test on paper. Some of that underground use is the problem I wrote about in [what to do about shadow AI](https://business.theaieducator.io/posts/shadow-ai-small-business). ## The Chore Rule The Chore Rule is how I suggest a small business gets reluctant staff using AI: start with the chore they would happily give away, never the craft they are proud of. Ask each reluctant person which part of their week they would drop tomorrow, put AI on that first, leave every judgment and the final say with them, and let what they save, not your enthusiasm, persuade the next person. It has three working parts. **The chore question.** Ask each reluctant person, one to one: "Which part of your week would you happily never do again?" Everyone has an answer, and it is rarely the answer an owner expects. For a senior estimator it might be typing up site notes. For an office manager it might be chasing the same three suppliers by email every Friday. For a sales lead it might be writing up call notes into the customer system. The chore is the job they already resent, so AI arrives as relief rather than as a verdict on their skill. **The veto.** Tell them plainly: "You can overrule anything it produces, and you never have to explain why." This answers the control shift the Harvard researchers describe. The person keeps the judgment that defines their expertise, and AI stays in the draft seat. It also puts your most experienced person where they are most useful: checking AI output they know better than anyone how to judge. **The proof passed along.** Do not measure logins. Measure hours handed back, in the person's own words, after two weeks. When your skeptic tells a colleague that the Friday supplier chase now takes ten minutes, that sentence does more than any launch email. Enthusiasm from the owner reads as pressure. Relief from a peer reads as evidence. The Chore Rule sits alongside the [One Job Rule](https://business.theaieducator.io/posts/how-should-a-small-business-start-using-ai), which picks the business's first AI job. The One Job Rule chooses the job for the business. The Chore Rule chooses the job for the person. ## What does this look like in a real small business? Take a hypothetical 25-person insurance brokerage. The owner bought business AI accounts for everyone in the spring. By autumn, the account handlers under 30 use them daily. The most experienced handler, who holds the firm's biggest clients, has opened the tool twice. The owner's first instinct is a demonstration: show the senior handler how AI can draft a renewal letter to a major client in thirty seconds. That is exactly the wrong move. The renewal letter is the senior handler's craft. The demonstration says, in effect, "a machine can do the thing you are proudest of." Under the Chore Rule, the owner asks the chore question instead. The answer is the notes after every client call: twenty minutes of typing, several times a day, that the handler has always resented. So AI goes there. The handler dictates a rough summary, AI turns it into a clean file note in the firm's format, and the handler reads it and corrects it before saving. The notes stay inside the firm's approved business account, and the handler can bin any note that is wrong. Two weeks later the owner asks one question: "What has that given you back?" If the answer is an hour a day, the handler will say so to the rest of the team without being asked. Nothing in that example requires a data team, a new system or a training day. It requires the owner to ask a question and listen to the answer. ## What I See When I Train Business Teams Across the business teams I train, the most skeptical person in the room is often the most experienced, and the objection is almost never "I can't learn this." I have trained more than 150,000 people across more than 30 countries. The quiet holdout usually has a good reason, and it tends to be one of three: they have seen AI make a confident mistake in their area, they are not sure what they are allowed to put into it, or they suspect that the part of the job they are good at is the part the business wants to automate. The first is answered by the veto. The second is answered by writing the rules down; Gallup's 68% against 47% is the case for doing it, and [a one-page AI policy](https://business.theaieducator.io/posts/does-a-small-business-need-an-ai-policy) is enough for most small firms. The third is answered by the Chore Rule, and only by it, because no amount of reassurance beats watching AI take away the job you hated while leaving the one you love. There is a quieter point too. Open skepticism is good news. The skeptic who asks "what happens when it is wrong?" is asking the question that protects the business. The one to worry about is the colleague who nods, says it is great and never opens it again. ## The Mistake I See Most Often The mistake is leading with the craft. Owners, understandably, want to show AI at its most impressive, and the most impressive demonstration is usually AI doing the skilled work: the proposal, the design, the client letter, the quote. To the person who does that work, the demonstration lands as a threat, and it is precisely the role compression the Harvard Business School research describes. They will be polite about it and then quietly carry on as before. The second most common mistake is counting usage. A dashboard of who logged in this week tells you who is willing to be seen using AI, and the Business.com and Harvard findings both suggest that is not the same as who finds it useful. If you must measure something, measure time handed back on a named job, reported by the person doing it. That is also the measure that tells you whether AI is [creating value or just activity](https://business.theaieducator.io/posts/why-small-businesses-are-not-getting-value-from-ai). ## What if someone still says no? Take the objection seriously, answer the part that can be answered, and be honest about the part that cannot. Privacy and security worries are often right, and Gallup found them behind the refusals of 43% of non-users. Answer them with rules and an approved business account, not with reassurance. Ethical objections deserve a hearing and a clear statement of what the business will not use AI for, such as deciding anything about a person on its own. "It can't help with my work" is sometimes true; run the chore question and see. Then be straight about the rest. If the business has decided that a job will be done with AI, say so, say why and give people the time and training to do it well. What does not work is pretending AI is optional while privately counting who uses it. People can tell. ## What should you do this week? You can do all of this in a 20-person business without a project plan. 1. **List the quiet ones.** Write down the people who have access to AI but rarely use it. Do not guess why. 2. **Ask the chore question, one to one.** "Which part of your week would you happily never do again?" Write down the answer in their words. 3. **Set up one chore with each of them.** Sit with them for twenty minutes and get AI doing a first draft of that chore in an approved business account. Keep it to one chore each. 4. **Say the veto out loud.** "You can overrule anything it produces, and you never have to explain why." 5. **Book the two-week check.** Put fifteen minutes in the diary with one question: "What has that given you back?" 6. **Stop counting logins.** If there is a usage league table, drop it. Ask people to share time saved instead. If you want the wider structure around this, the [Four Rungs of AI training](https://business.theaieducator.io/posts/ai-training-for-small-business-teams) explain why the chore also has to come with permission, practice and a return visit. ## Where to go next If your team has AI but the people who matter most are not using it, that is exactly what my [AI training for teams](https://theaieducator.io/ai-training-for-business?utm_source=business.theaieducator.io&utm_medium=referral&utm_campaign=how-to-get-employees-to-use-ai) is built for: practical sessions on your own jobs, skeptics included. If you want to know where your business stands first, take the free [Workplace AI Readiness Check](https://theaieducator.io/workplace-ai-readiness?utm_source=business.theaieducator.io&utm_medium=referral&utm_campaign=how-to-get-employees-to-use-ai): 21 questions, about eight minutes, scored across seven dimensions. ## Sources and further reading - [AI in the Workplace: What Separates Adopters and Holdouts](https://www.gallup.com/workplace/704252/workplace-separates-adopters-holdouts.aspx), Gallup, April 12, 2026 (survey of 23,717 US employed adults, February 4 to 19, 2026). - [2026 Small Business AI Outlook Report](https://www.business.com/articles/ai-usage-smb-workplace-study/), Business.com, January 20, 2026 (survey of 1,009 people at US businesses with 2 to 250 employees). - [Why Employees Resist AI, and How Companies Can Win Them Over](https://www.library.hbs.edu/working-knowledge/why-employees-resist-ai-and-how-companies-can-win-them-over), Harvard Business School Working Knowledge, June 24, 2026, on research by Das Narayandas and Shunyuan Zhang. - [WRITER Survey Finds 60% of Companies Plan to Lay Off Employees Who Won't Adopt AI](https://writer.com/blog/enterprise-ai-adoption-survey-results-press-release/), WRITER with Workplace Intelligence, April 7, 2026 (2,400 respondents, fieldwork December 17, 2025 to January 25, 2026). *[Dan Fitzpatrick](https://theaieducator.io/about) is The AI Educator, a Forbes contributor, international keynote speaker and bestselling author who helps businesses use AI well, safely and every day.* ## Frequently asked questions ### How do I get my employees to use AI? Start with a chore, not a mandate. Ask each person which part of their week they would happily never do again, set AI up on that job with them, let them overrule anything it produces, and check after two weeks what time it has given back. Relief spreads faster than instructions. ### Why do employees resist using AI at work? Mostly because they prefer their current way of working. In Gallup's 2026 survey, 46% of US workers who do not use AI gave that reason, ahead of ethical objections, privacy and security concerns, and a belief that AI cannot help with their work. ### Should I make AI use mandatory for my staff? Not as a first move. Pressure tends to push resistance out of sight, with people appearing to use AI without really relying on it. If the business decides a job will be done with AI, say so openly, explain why, and give people time and training to do it well. ### How do I know if my team is really using AI? Ask what it has given back, not who logged in. Usage counts show who is willing to be seen using AI. Time saved on a named job, reported by the person doing it, shows whether AI is changing the work. ### What should I do about an experienced employee who refuses to use AI? Ask them which chore they would drop tomorrow and put AI there first, leaving the skilled part of their job untouched. Tell them they can overrule anything AI produces. Experienced staff are often the best checkers of AI output, so give them that role. ### Does AI training fix staff reluctance? Only if it lands on a real job the person wants help with. Training on general features rarely changes behavior on its own. Pair it with a named chore for each person, clear rules on what is allowed, and a follow-up a week or two later. --- Source: [How to Get Skeptical Employees to Use AI in a Small Business](https://business.theaieducator.io/posts/how-to-get-employees-to-use-ai) Publisher: [The AI Educator](https://theaieducator.io)