AI FOMO is not a strategy: is it a race to the bottom?
Almost every conversation we engage in today centres on one thing: AI.
The fear of falling behind is creating anxiety in businesses of all shapes and sizes. In many cases, it’s propelling them into making technology choices with scant regard for strategy. In others, they’re stuck. Paralysed by the vast array in front of them, and the pace with which it changes. Neither is a particularly happy place.
The race itself isn’t necessarily the problem. But knowing which race you’re in – and why – is crucial.
Some AI adoption will be a baseline requirement to deliver efficiencies that keep you on a par.
Some will be a race to the bottom, where relentless cost cutting risks degrading differentiation, service or value in the process.
And if you join the race without being clear about what you’re trying to achieve, the consequences may not just be wasted investment. Failed initiatives, disengaged staff and ticked off customers are well-documented risks.
Less well-understood – or talked about – are the risks that emerge around AI itself: what happens to your information, what you become dependent on, what knowledge you lose, and whether your governance can keep pace.
And then there’s the race you want to win: creating advantage.
It is here where deliberate commercial choices are non-negotiable. They can be framed by the following key questions:
What are we actually trying to achieve?
What does AI need to work here?
What do we need to stay in control?
What needs to change around it?
What happens when things go wrong?
Want to answer these for your business? Read our practical guide →
What are we actually trying to achieve?
One of the most frequent questions we get asked is “what should we be doing about AI?”. The trouble is, there is no single answer. Only more questions. Beginning very clearly with what you want to achieve.
The answer has to start with what you know best: your business and your competitive landscape. Efficiencies, while appealing, are far less interesting than effectiveness – not just doing things faster or cheaper but better to deliver value.
So where could doing something better genuinely change your outcome?
Could you serve customers in a way you couldn't before? Make better use of knowledge you already have? Give people information they couldn't previously access or act on? Remove a constraint that has always limited what you could do?
That doesn't mean ignoring efficiency. There will be places where using AI to do the same thing faster or at lower cost is commercially important – particularly as those efficiencies become the minimum expected to keep pace. But that's a different strategic choice from using AI to create an advantage.
And neither starts with the technology. It starts with being clear about the business outcome, how much it matters and whether AI is the best way to achieve it.
What does AI need to work here?
AI cannot function without data. Though data alone isn’t enough. Technology, processes, skills and adoption all matter. But without well-ordered, clean and accessible data, the chances of AI delivering the outcomes you want diminish.
And this is where AI ambition can crash into reality.
Most businesses weren't designed around the data demands of AI. Information has accumulated over years, often across different systems, teams and suppliers, with different definitions, owners and levels of quality. People know where to find things, which numbers to trust and which workarounds to use. Until you ask AI to make sense of it.
Suddenly, inconsistencies that were manageable when a human knew the context become much more significant. If customer records don't match, information is incomplete, permissions aren't clear or nobody is quite sure which version is authoritative, adding AI doesn't resolve the problem. In fact, it often amplifies it.
That doesn't mean you need perfect data before you can do anything with AI. You probably never will. It means understanding what data a particular application depends on, whether it's good enough for the job you're asking it to do, and what would need to change if it isn't.
You don't need to solve every possible problem before you start. But you do need to understand the foundations the outcome you're pursuing depends upon.
What do we need to stay in control?
Policy is only part of the answer. Policy isn’t process.
Consider:
What information is going where?
People may be putting commercially sensitive information, customer data or IP into tools. Do you know what information is being shared, where it goes, how it is used and who has access to it?
What are you becoming dependent on?
A model, vendor, feature or automated process can quietly become operationally important. What happens if it changes, disappears, becomes more expensive or starts behaving differently?
Who owns the choice and outcome?
Not just who bought the software. Who decides appropriate use? Who understands the consequences? Who is accountable when an AI-supported process produces something consequential?
AI can also enter your business without a conscious decision to introduce it. Increasingly, AI functionality is being built into technology you already use. A useful new feature can also change how information is handled and who else is involved in providing the service.
What’s more, the pace of governance rarely mirrors the pace of change. The answer isn’t necessarily more control. Indeed, excessive control can simply drive usage underground. The objective is proportionate governance that lets the business exploit the opportunity without losing sight of what it's doing.
What needs to change around it?
Introducing AI inevitably changes more than the technology. It can change roles, responsibilities, workflows and the skills people need. It can also change the knowledge and judgement you need people to retain.
As AI takes on work that people previously had to understand or do themselves, it's easy for capability to disappear almost unnoticed. That may not matter when everything works as expected. It matters considerably more when it doesn't.
So think beyond training people to use the tool. What will people do differently? What do they still need to know? Where does human judgement remain important? And how will you know whether the change is delivering the outcome you wanted?
What happens when things go wrong?
No matter how good the technology, things will go wrong. AI will produce the wrong answer. Suppliers will have problems. People will use tools in ways you didn't anticipate.
The important question is whether you are prepared to recognise and respond when they do. Can people spot when something isn't right? Do they know when to intervene and who to escalate to? Can you still operate if a system or supplier becomes unavailable? What happens if data or information becomes compromised?
Run your own race
The pressure to do something about AI isn't going away. Nor should it. The opportunities are significant, and standing still carries risks of its own.
But speed isn't the same as progress.
You don't need to know exactly where AI will take your business. You do need to be clear about why you're using it, what it depends upon, what you're changing in the process and how much risk you're prepared to accept.
The businesses that create advantage won't necessarily be those that adopt AI fastest. They'll be the ones that know which race they’re in – and why.
Thinking about where AI fits into your business? Work through our practical guide before you invest or get in touch.