Find the balance between AI adoption and your digital strategy

Every board meeting has it on the agenda now. AI. What are we doing with it? Where are we deploying it? What are our competitors doing that we aren't?

The pressure to act is real. The fear of falling behind is real. But the decisions being made under that pressure are often the wrong ones.

I've spent nearly four decades watching technology cycles play out. The pattern is familiar. A new capability emerges, the hype builds, businesses rush to adopt it without a clear reason to, and then they spend the following years untangling the mess. AI is following the same curve. The difference is the speed, the cost, and the environmental consequences are all significantly higher.

The problem with chasing the trend

Most organisations adopting AI right now are doing so reactively.

A competitor announces something. A vendor demo looks compelling. A board member reads an article. Suddenly there's pressure to ship something, anything, with AI in it.

What follows is uncontrolled experimentation. Different teams spinning up different tools. No shared understanding of what's appropriate or how data should be handled. No clear line between the AI activity and the wider digital strategy.

The costs compound quietly. Compute bills arrive. Data governance questions surface, and somewhere in the middle of it all, someone asks:

what problem were we actually solving? That's not an AI problem. It's a strategy problem.

Strategy before scale

AI should be treated as a strategic capability, not a novelty to be deployed and figured out later.

That means doing the thinking before doing the spending. What outcomes are you trying to achieve? Which parts of your business would genuinely benefit from AI assisted decisions or automation? Where does AI add a layer of complexity that a simpler process could handle just as well? These aren't difficult questions, but most businesses skip them in the rush to look current.

When you're clear on outcomes first, the AI decisions that follow are significantly sharper. You deploy where it matters, not everywhere you can.

Define your guardrails before you deploy

The businesses getting AI right have one thing in common: they defined their guardrails before they started.

An AI usage policy isn't bureaucracy. It's clarity. It answers the questions that will otherwise trip you up later: What data can be passed to external models? Who's accountable when an AI generated output causes a problem? What does appropriate use look like in your context?

Without that clarity, you're not deploying AI strategically. You're hoping the individuals making daily decisions happen to get it right.

The same principle applies to model selection. There's a persistent assumption that the biggest, most powerful model is always the right choice. It isn't.

A large language model optimised for complex reasoning costs more to run, consumes more energy, and often delivers no better result for routine tasks than a smaller, well-chosen model. Strategic orchestration, knowing which model to use for which task, is where real performance and cost control come from.

This matters commercially. It also matters environmentally.

The environmental cost nobody's talking about

Running AI models at scale has a carbon footprint that most organisations aren't measuring.

The compute required to train and serve large models is significant. When you multiply uncontrolled AI experimentation across dozens of teams, the environmental impact adds up fast and it's rarely appearing in any sustainability report.

Scaling responsibly isn't idealism. It's competitive advantage.

When you select models deliberately, manage token usage with discipline, and choose providers who are transparent about their energy sourcing, you reduce costs and reduce environmental impact at the same time. Those two things aren't in tension.

The businesses that understand this are building AI capability that they can account for to investors, customers, and regulators. The ones ignoring it are storing up a problem.

AI should support your digital strategy, not compete with it

This is the point that often gets lost.

Your digital strategy exists for a reason. It reflects the direction of your business, the experiences you're building for customers, the systems you're investing in for the long term.

When AI is adopted without reference to that strategy, it starts to pull in a different direction. Priorities fragment. Resources get consumed by AI experimentation that doesn't connect to anything that matters. The digital roadmap loses coherence. AI adopted deliberately, with clear outcomes, defined guardrails, and thoughtful model selection, strengthens your digital strategy rather than competing with it. It fills genuine capability gaps. It improves the performance of systems you've already invested in. It delivers measurable returns.

That's a very different outcome from deploying AI because the pressure to do something became too loud to ignore.

Let's wrap this up

AI can create real competitive advantage. It can also create unnecessary complexity, rising cost, and environmental impact that nobody planned for.

The difference comes down to whether you treat it as a strategic capability or a trend to be chased.

Before you scale, define your guardrails. Build your AI usage policy. Set your model selection criteria. Be deliberate about sustainability. Then adopt AI in ways that reinforce your digital direction rather than fragment it.

If you're working through what that looks like for your business, I'm happy to talk it through with you.