AI integration is now a judgement problem

Most leadership teams no longer need convincing about the relevance or potential of AI. Since the release of ChatGPT in 2022, industry discussion has shifted away from whether it matters, towards what should actually be done with it inside an organisation. 

That shift is important because it changes the nature of the challenge before leadership teams. AI is no longer just a strategic concept to be explored, but a set of decisions that need to be made under real constraints of time, capacity and organisational focus. 

This is where AI stops being a strategy slide and becomes a judgement test.

The risk at one end is inflation. Every team claims relevance. Every use case sounds important. The conversation fills with possibility, but very little is prioritised. The firm appears ambitious, yet nobody is forced to decide where value is actually most likely to come from.

The risk at the other end is hesitation. Governance concerns multiply, pilots continue and the organisation talks intelligently about AI while making very few decisions that change how work is actually done.

Neither is especially strong leadership. One is overreach. The other is drift.

The real executive task is narrower and more demanding: leaders need to make disciplined choices about where AI will genuinely improve decision quality, productivity or client experience, and where it will not. That means being explicit about how work is reconfigured in practice, distinguishing between what should be automated, what should be augmented and what must remain firmly human.

It also requires a clear sense of organisational capacity. AI adoption is not just about what is technically possible, but about what the firm can realistically absorb without creating fragmentation, confusion or competing priorities elsewhere. Just as importantly, responsibility for these decisions needs to be unambiguous, so that intent translates into execution rather than remaining at the level of experimentation.

That is why this is no longer just a technology issue. It is a leadership one. 

In many firms, the biggest constraint is not the software. It is the inability of senior people to make a small number of clear decisions and hold the line around them.

Good judgement here requires restraint as much as ambition. It means being willing to say no to interesting things so, instead, a few important things can be done. And done properly. It means being honest about the state of the organisation. For instance, asking questions such as do we really have the workflow discipline, data quality and management bandwidth to support what we are proposing? If not, the answer may still be yes in principle, but not yet in practice.

Overall, the firms that benefit most from AI are unlikely to be the ones with the noisiest language around it. They are more likely to be the ones whose leaders can decide clearly, sequence sensibly and turn a broad opportunity into a manageable set of real changes.

That is what executive judgement looks like here.