Five Questions Business Leaders Actually Ask: What AI Workshops Reveal About Real Adoption

“Is It Safe to Use with My Customer Data?”

This is always the first question. Always.

In a recent workshop, a director of an estate agency with hundreds of thousands of customer records put it bluntly: he wanted to understand AI, he could see the potential, but the idea of feeding client data into a system he did not fully understand made him deeply uncomfortable.

He is right to be cautious. And the answer is more nuanced than most AI enthusiasts care to admit.

The short version: consumer AI accounts (the free versions most people start with) offer limited privacy guarantees. You often cannot control whether your data is used for training. Business and enterprise accounts from the major providers — Anthropic, OpenAI, Google — contractually commit to not training on your data, encrypting it, and not allowing human access.

But “contractually commit” is not the same as “guarantee.” The only true guarantee is running AI models locally on your own hardware, which is becoming more feasible but still requires significant investment.

The first wins come from internal documents, reporting, email drafting, and strategic planning. None of that requires customer records.

“I Use the Free Version. Is That Not Good Enough?”

This is where the room always goes quiet.

The uncomfortable truth is that free AI models are dramatically worse than paid ones. Not slightly worse. Dramatically worse. They will confidently give you incorrect answers to straightforward questions, fabricate data in financial summaries, and produce marketing copy that reads like it was written by a committee that has never met your customers.

The business model is simple: it costs these companies significant money to run their best models. If they give you the best for free, they lose money on every interaction. So the free tier runs cheaper, faster, less capable models — and most users have no idea this is happening.

The investment is modest. For most business professionals, a subscription of twenty to thirty pounds per month unlocks a fundamentally different experience. If AI saves you even two hours per week — and it will save far more than that — the return on investment is immediate.

“How Do I Actually Set This Up for My Business?”

This is the gap that nobody talks about. Business professionals understand the potential. They have read the articles. They have seen the demonstrations. But when they sit down at their desk on Monday morning, they do not know where to start.

The answer is simpler than most people expect: projects.

Every major AI platform now supports what they call projects — dedicated workspaces where you set instructions, upload reference files, and define how the AI should behave for a specific workflow. Think of it as briefing a new team member. You tell them your standards, give them the relevant documents, and explain how you want things done.

This is consistently the single most impactful thing business professionals can do in their first week with AI. Not building chatbots. Not automating customer service. Just setting up three or four projects that match their actual daily workflows.

“Will This Replace My Staff?”

A publishing company owner raised this with unusual honesty: he could see how AI would help, but he had a moral hesitation about where it might lead for his team of fourteen freelancers.

This question deserves a direct answer, and the answer is different for small businesses than for large corporations.

Large companies are already laying off thousands. Block, the parent company of Square, recently eliminated over four thousand roles and explicitly cited AI as the reason. Meta has signalled similar intentions. For corporations managing thousands of employees, AI represents an opportunity to cut costs at scale.

For small businesses, the calculus is different. A team of five or ten people is not carrying the kind of organisational overhead that AI eliminates in large corporations. What AI eliminates in small businesses is not people — it is the bottleneck that prevents those people from doing more valuable work.

What I have seen consistently is that small businesses that adopt AI do not shrink. They grow. Often without needing to hire the additional person they thought they needed. The revenue comes first. The hiring decision — if it is even necessary — comes later, from a position of confidence rather than risk.

“Where Do I Actually Start?”

This is the unknown unknowns problem, and it is the real barrier to AI adoption. Not cost. Not privacy. Not technical skill. It is simply not knowing what is possible.

The starting point is always the same: look at your week. Identify the three tasks you dislike the most — the ones that drain your energy, take too long, and feel like they should be easier. Those are your first AI projects.

For most business professionals, this list includes some combination of: email drafting, report writing, data summarisation, meeting preparation, marketing copy, financial analysis, and document formatting.

Pick one. Set up a project. Give the AI your context — your style, your standards, your reference documents. Try it for a week. Then pick the next one.

The transformation is not instantaneous. It is iterative. Each week you find another workflow that AI can handle. Each month your capacity expands. Within three months, most business professionals describe a fundamentally different relationship with their workload.

Not because they are working less. But because the cognitive load has shifted. The draining, repetitive work gets handled. The strategic, creative, relationship-building work gets the time it deserves.

And that is where the real value lies — not in AI doing your job, but in AI clearing the path so you can do the part of your job that only you can do.