Your business is probably already using AI; at least, that's how it feels. Your people use ChatGPT or similar tools every day, and the productivity gains are real. But that AI sits with the individual, not the business. When a staff member leaves, the capability walks out with them. The business never owned it, can't audit how it was used, and can't reproduce the result without that person.
There are two fundamentally different kinds of AI activity going on in most businesses right now, and they get confused every day.
Personal AI
Belongs to the individual
Leaves when they do
Can't be audited or reproduced
Real productivity, no durable value
Worth encouraging. Not an asset.
Business AI
Owned and operated by the organisation
Built deliberately, documented to run on
Governed, audited, with named accountability
Scales beyond any one person
Owned, governed, durable. The strategic investment.
Personal AI is worth encouraging. Business AI is the only kind that becomes a strategic investment: systems the organisation itself owns, operates and is accountable for, with data, access and permissions defined to the business's own risk appetite. If your business uses AI only through your people's personal tools, it is not yet using AI in ways that create lasting value.
Every business will have an AI strategy
The only question is whether it will be a deliberate one, designed by the leadership team for the business's competitive position, or a reactive one, assembled in response to what competitors do first.
A deliberate AI strategy starts from the business, not the technology. It looks at what the business already has, where customers are pulling, where the friction sits, what the competitive landscape rewards, and what the organisation can realistically absorb. Then it ranks the opportunities by what they would cost, what they would return, and how much each would bend the business out of shape. The opportunities usually fall into recognisable shapes: new value from what you've already invested in, internal process automation, customer retention, and overhead reduction.
Where AI shows up: the opportunity radar
The radar is a useful map of where AI can appear across a business, and of where the everyday moves end and the game-changing ones begin. Most businesses find their current AI activity sits on the left, in Process AI: useful, safer, and the smaller prize. The strategic question is whether the right next move is more depth there, or a shift to the right, to Transformational AI that changes what you sell, how you serve, and how easily customers leave.
Process AI automates existing work. Transformational AI changes how the business creates value. Both are useful; only one creates a position competitors find hard to reverse.
Three principles that separate durable advantage from noise
There is no magic
AI amplifies work that humans have properly scoped. The discipline is defining the problem precisely: what success looks like, where the boundaries are, what evidence will validate the answer.
Trust, but verify
AI executes; humans verify. Where AI is in the production loop, human review, audit and accountability boundaries are designed in from day one, never retrofitted after a problem appears.
Clear human accountability
Every AI-touched decision has a named human responsible for the output, an explicit sign-off path and an audit trail. Governance is designed in before the board, the regulator or a customer asks.
Where to start
The right place is rarely the obvious one. It isn't always the function with the loudest pain or the most enthusiastic sponsor. It's the place where AI compounds what you already have, your customers, your operating data, your know-how, your brand, and where the move is hardest for a competitor to copy or reverse.
Three phases, adoption, productionisation and governance, and within them the difference between Process AI and Transformational AI. Finding the right starting point is the first job of an AI strategy.
Faster and cheaper than you'd expect
What used to take a cross-functional team the better part of a year can now be prototyped in weeks and in production in months, with a senior generalist working with AI as the building tool. For a mid-sized business without a development team, and without the appetite to hire one, that is the difference between "AI is something we'll get to eventually" and "AI is something we can act on this quarter."
The point
Lean on AI, but don't delegate to it. Leadership remains accountable for what the business does; AI amplifies what leadership decides. Strategy first; the rest only if and when the strategy says it should.