Your Employees Are Afraid AI Will Replace Them. Here Is What a Small Employer Can Say Without Promising Too Much

Tell your team what you are using AI for, what you do not know yet, and what you will commit to: notice, paid training time and a person to ask. Do not promise that nobody will lose a job. In California a casual promise of job security can be argued as an implied contract, and you cannot keep it if the business changes.

I am Mac Malbrough. I hold a J.D. and an MS-HRM, I am not an attorney, and I write for owners and managers who can see the worry on their team’s faces and are not sure what they are allowed to say. People have started calling this FOBO, the fear of becoming obsolete. The label is new. The conversation you need to have is old: how do you talk straight to people when you cannot guarantee their future?

How common is the fear?

Common enough to plan for. A Pew Research Center survey of 5,273 employed U.S. adults, fielded October 7 to 13, 2024, found that 52% of workers were worried about the future use of AI in the workplace, 36% were hopeful and 33% felt overwhelmed. Only 16% said some of their own work was being done with AI at the time, so most of that worry was about something that had not happened to them yet.

A more recent snapshot comes from the UK. A June 2026 survey of more than 5,000 workers by the GMB union, reported by Workplace Journal, found that 48% were concerned about AI’s impact on their jobs and 58% believed AI would eliminate jobs in their workplace. It was not an American sample, and a union ran it, so I treat it as a direction, not a benchmark. The two surveys point the same way: the fear is widespread, and it runs ahead of the technology.

What does the fear cost a small team?

It shows up as quiet behavior rather than complaints. People stop volunteering ideas that would make their own task faster. They hide how they use the tools, or they refuse to try them. The best ones, who have options, update their resumes. You do not need a number for this to be a problem, because in a team of 25 a single resignation is a visible loss.

Silence makes it worse. When an employer says nothing, employees fill the gap with the worst version of the story. When an employer says something vague like “AI will not replace you, it will help you,” people hear a line from a press release and trust you less.

What should you say in the first conversation?

Keep it to five minutes at a team meeting, and say it in your own words. Here is the structure I would use:

• “Here is what we are using AI for right now: [name the actual tools and tasks].”
• “Here is what we are not using it for: [hiring decisions, discipline, whatever is true].”
• “Here is what I do not know: how these tools will change our work in a year.”
• “Here is what I will commit to: if a role is going to change, you will hear it from me first, and you will get paid time to learn the new way.”
• “If you have a concern, bring it to [named person], and I will answer within a week.”

Fill the brackets with facts. The “what I do not know” line is the one that builds trust. People forgive an employer who says “I am not sure.” They do not forgive one who says “you are safe” and then makes a cut.

Why should you not promise job security?

California presumes employment is at will under Labor Code 2922, which means either side can end it at any time, with or without a reason. Courts have recognized that employer statements, handbook language and a pattern of assurances can support an argument that the employment is not at will. A manager who says “no one is losing a job to AI on my watch” in writing hands a former employee a quotation.

The better wording describes your process rather than the outcome. “We will give at least 30 days’ notice before a role changes materially” is a commitment you control. “Nobody will be laid off” is a prediction about the future of your business. If you do not already have it, your handbook’s at-will paragraph should say that your policies are not a contract, and my California employee handbook template includes that language.

How do you build AI skills without scaring people more?

Offer training as an investment in the person, not a requirement attached to a threat. A few choices help:

• Do it on the clock. Time in training you require is generally time worked. Making people learn on their own time, while they fear the same tool will replace them, is a bad trade.
• Start with their own tasks. The best first lesson is “show me how to use this on a report you already write.” It makes the tool theirs rather than management’s.
• Let people opt in to the first pilot. Volunteers report honestly, and their results persuade peers better than a memo does.
• Ask what they would automate. The people closest to the work know which tasks are tedious. That puts employees on the side of the tool instead of the receiving end.

What about monitoring and honesty about the tools?

The GMB survey found that 48% of respondents believed employers use AI to monitor their own or colleagues’ activity. If you use a tool that records, scores or tracks anything about your employees, say so in plain language. Hidden monitoring turns a worry about the future into a worry about the present. My piece on AI notetakers in HR meetings covers California’s recording consent rule, which matters here.

What should managers do between meetings?

The team meeting starts the conversation. Managers keep it going in one-to-ones. Ask each person what they use the tools for, what worries them about the next year, and what they would like to learn. Write down what you hear and follow up on one thing within two weeks, because a promise kept on a small item earns more trust than a speech.

What should you do this week?

Write the five-line script above with your real tools and your real commitments, and test it on one trusted manager. Check that nothing in it promises an outcome you cannot control. Then book the five minutes at your next team meeting and name the person who will take questions.

If you want a second opinion on the wording, or a handbook paragraph for how your company uses AI, I can help through fractional HR. My piece on managers and retention explains why this conversation is often the one that keeps your best people.

For a way to place your own team, see the four stages of AI use on an HR team.

Related reading: a green, yellow and red map of where AI helps HR.

For a related angle, see the manager skill of AI judgment.

Sources

• Pew Research Center, U.S. workers are more worried than hopeful about future AI use in the workplace, February 25, 2025, read October 31, 2026
• Workplace Journal, Nearly half of workers fear AI could replace their jobs, June 9, 2026, read October 31, 2026
• Labor Code section 2922, read October 31, 2026

McKinley holds a J.D. but is not a licensed attorney. Articles here are general information, not legal advice. For your specific situation, talk with an employment attorney.