AI at Work Isn’t an IT Problem - It’s a People Problem

AI is already being used in your workplace, whether you’ve formally approved it or not. The real risk isn’t that people are using it. It’s that they may be using it without any guidance.

Here’s how to put some simple, practical and human guardrails in place.

Ask most business owners whether their team is using AI at work and the answer is often, “Not really.” But look a little closer and you’ll usually find a different story.

Someone is using it to tidy up a client email. Someone else is asking it to summarise a long report, draft a job advertisement or make sense of a spreadsheet.

AI has quietly become part of how work gets done. In many small businesses, though, no one has really talked about it.

That’s the part we need to pay attention to.

The risk is rarely that people are using AI. The greater risk is that they’re using it without knowing what’s okay, what isn’t and where things could go wrong.

AI has already walked through the door

It’s easy to think of AI as a conversation for the future. The reality is that it’s already being used in most workplaces, often without the business ever formally deciding that it should be.

People reach for these tools because they’re busy and the tools can be genuinely helpful. Ignoring that doesn’t stop it from happening. It simply means it’s happening without visibility or agreed expectations.

The first step isn’t necessarily to ban AI or lock everything down. It’s to acknowledge that it’s here and get curious about how your team is already using it.

This is a people issue, not just a technology issue

It can be tempting to hand AI over to whoever looks after your technology and consider the job done. But many of the most important questions aren’t technical ones.

What information is safe to enter into an AI tool? Who is responsible if the information it produces is wrong? How do we protect our clients, employees and commercially sensitive information? Where must human judgement remain firmly in charge?

These are leadership, culture, risk and accountability questions. That’s why AI needs to sit on the HR and leadership agenda…not only on the IT to-do list.

The risks worth talking about

You don’t need to be an AI expert to understand where the risks usually arise:

●     Confidentiality and privacy: Client information, employee details and commercially sensitive data should never be entered into public AI tools.

●     Accuracy: AI can sound very confident, even when the answer is incomplete or simply wrong. If no one checks the output, its mistake can quickly become your mistake.

●     Accountability: “The system told me to” isn’t an acceptable explanation. A person must still be responsible for reviewing the work and owning the outcome.

●     Fairness: Anything affecting recruitment, remuneration, performance or employment decisions needs careful human judgement. These are not areas where we should rely on an unchecked automated shortcut.

A one-page guide can make a real difference

The good news is that managing AI well doesn’t require a major technology project or a significant investment.

For many small businesses, a short, plain-English AI guide is a sensible place to start. It should answer a few practical questions:

●     Which AI tools can employees use?

●     What information must never be entered?

●     How should AI-generated work be checked?

●     When does a person need to review or approve the outcome?

●     Who should employees speak to if they’re unsure?

Keep it simple, make it relevant to the work your people actually do and talk your team through it. One clear page that people understand and use is far more valuable than a lengthy policy that no one reads.

Talk with your team - don’t police them

The goal should be guidance, not surveillance.

If employees think they’ll be criticised or punished for admitting they use AI, they’ll simply stop talking about it. That means the business loses visibility and the opportunity to make its use safer.

Ask your team how they’re already using AI. What is helping? What feels unclear? Where do they see risks? What tasks could potentially be done more efficiently?

Build the guardrails with them. You’ll reduce risk, but you may also uncover some genuinely useful ideas for saving time and improving the quality of work.

A note on your obligations

AI use also needs to be considered alongside your existing privacy, employment and work health and safety obligations.

When automated tools begin influencing how work is allocated, assessed or monitored—or start shaping decisions about recruitment, performance or employment—the risks become much broader than productivity.

Businesses still need to consider the impact on people, consult where appropriate and keep meaningful human oversight in place. Technology may support a decision, but it shouldn’t remove human responsibility for it.

The bottom line

AI isn’t going away. Used well, it can save small businesses time and improve the quality of everyday work. But convenience without clear guardrails is where problems begin.

You don’t need to understand every part of the technology to lead this well. Start with an open conversation and a simple set of expectations that everyone understands.

AI at work is ultimately a people issue and it’s worth getting ahead of it before a problem forces the conversation.

If you’d like some help putting a practical, plain-English AI policy in place, this is exactly the kind of work I help business owners navigate.