Ask the owner of a 40-person business what the team thinks about AI and there is often a pause. It is not that nobody has a view. It is that nobody has said anything first, so everyone is guessing.

Littler, a large US employment law firm, published its 2026 Labor Survey Report on 29 September. It asked 665 in-house lawyers, HR professionals and business executives how ready they feel for the employee relations side of AI. Only 9% said they feel "very prepared" to manage the labour and employee relations issues AI adoption may create. Just 17% said they communicate proactively and consistently with employees who are worried about AI's impact.

Some honest caveats. Ninety-three percent of respondents are US-based, the survey is framed around labour relations, and it asks employers, not their staff. It is not a UK statistic and should not be read as one. What it offers is a pattern that most small-business owners will recognise: the people in charge know a conversation is needed, and most have not started it.

Jonathan Levine, who co-chairs Littler's Labor Management Relations Practice Group, put the underlying problem in one line: "Some employers treat AI as an IT project when it's really a workplace change project."

That is worth taking seriously, because it changes what the first step looks like. An IT project begins with choosing a tool. A workplace change begins with telling people what is going on. In a 40-person firm the second is far cheaper, and it tends to decide whether the first ever gets used.

Silence has a cost. When a leader says nothing about AI, staff fill the gap themselves. Some assume the worst about their role. Some quietly use personal accounts and hope nobody asks. Others avoid the tools altogether because they are not sure whether using them is welcome. None of this reflects badly on the people involved. It is what happens when information is missing.

The good news for a smaller business is that the fix is modest. It is a conversation, not a programme. A short team meeting is enough to cover the ground, and the content does not need to be polished. It needs to be plain and true.

It helps to cover a handful of things. What AI is already being used for in the business, including by the owner. What the business is trying out next and why. What is not changing, said only where it is genuinely true. What is still undecided, which is usually a lot. And where people can ask questions without it counting against them.

The "what is not changing" part deserves care. It is tempting to promise that nobody's job is at risk. If that is not something the owner can honestly guarantee, it does not belong in the room. A more useful sentence is often "we are not planning changes to anyone's role as part of this, and if that ever changes, you will hear it from me first." That is a smaller promise and a far more believable one.

AI can help with the preparation. A leader can give Claude their rough notes and ask for a one-page, plain-English summary of where the business stands on AI, then edit it until it sounds like them. ChatGPT or Gemini will do a similar job. What matters is that the final words are the leader's own, and that the page invites questions rather than closing them down.

Then repeat it. Fifteen minutes each month, with the same few questions, does more than one big announcement. People do not need certainty from their leaders. They need to know that when something changes, someone will tell them straight.

There is a quieter benefit too. A team that has heard a calm, honest account of what AI means for the business is much more likely to try it on real work and to say when something is not working. The conversation is the first piece of adoption, not the preamble to it.

Go Wisely works with small and mid-sized businesses on exactly this: the human and organisational side of adopting AI, where the hard part is rarely the software. The tools will keep changing every few weeks. The habit of explaining plainly what is happening to the people affected by it will keep being useful.

Go wisely.