AI makes collective judgement in leadership teams even more important
Investment in artificial intelligence is accelerating. Organisations are developing AI strategies, automating workflows and expecting higher productivity and better decisions. Much of the enthusiasm is justified. AI can analyse large volumes of data, summarise complex material, develop scenarios and prepare decision papers in a fraction of the time previously required.
But amid the fascination, we are overlooking a more fundamental shift. As analysis becomes faster and cheaper, the value of leadership shifts. What matters is no longer simply the ability to gather information or produce well-founded answers. It is the ability to interpret those answers, challenge them, prioritise between them and translate them into collective action.
When intelligence becomes cheap, judgement becomes valuable
Organisations rarely need more answers
Most organisations already have more analyses, strategies and decision papers than they can put into practice. The problem is rarely that leadership teams do not know enough.
The problem arises when different parts of the organisation interpret the same situation differently, prioritise according to their own interests and subsequently act in different directions.
A strategy can be clear on paper yet remain unclear in practice. Leaders can agree on the overall objectives while fundamentally disagreeing about what should actually be prioritised, which resources should be reallocated and what the organisation should stop doing.
AI does not solve this challenge. On the contrary, it may make it more visible.
When every executive can produce well-supported arguments, analyses and recommendations, it becomes easier to make informed decisions. But it also becomes easier to justify one's own position. Marketing can demonstrate the need for greater investment in the market. Finance can demonstrate the need for cost reductions. HR can demonstrate the need for skills development. Operations can demonstrate the need for greater stability.
Everyone may have valid arguments. The organisation still cannot do everything. This is not an information problem. It is a prioritisation and collaboration problem.
AI amplifies existing patterns
We often hear warnings that AI could create new silos. That is only partly true. AI does not necessarily create the silos. It amplifies the organisational patterns that already exist.
In an organisation with shared direction and strong coordination, AI can improve the basis for decisions, reduce unnecessary work and accelerate organisational learning.
In a fragmented organisation, the same technology can generate more local analyses, greater output and increased confidence in each function's own view of reality. The organisation becomes faster, but not necessarily more coherent. It may simply move faster in different directions.
AI therefore becomes a stress test for the leadership team. It reveals whether the leadership team functions as the organisation's collective decision-making system or as a group of competent leaders who meet, exchange information and then return to their respective areas.
Technology is not neutral in practice. It is applied through existing structures, incentives, relationships and power dynamics. If leaders continue to be measured and rewarded primarily on the performance of their own areas, AI will probably be used first and foremost to optimise those areas.
You cannot compensate for a fragmented leadership system with better technology.
The greatest risk is artificial agreement
The debate about AI risks often focuses on hallucinations, bias and factual errors. These are genuine concerns. AI-generated analyses must be checked, and responsibility for decisions cannot be delegated to a model.
But the organisational risk extends beyond individual errors. The risk is that convincing language is mistaken for sound judgement.
AI can produce coherent arguments with considerable linguistic confidence. An answer may appear complete without its underlying assumptions having been examined. An extensive decision paper may look thorough even when important perspectives are missing.
Leadership teams therefore risk using AI to close discussions that the technology should have helped them open.
If leaders arrive with their own AI-generated conclusions but fail to examine the data, interests and assumptions behind them, leadership meetings become an exchange of plausible answers rather than a collective examination of the organisation's reality.
Professional disagreement becomes a form of quality control
The better AI becomes at producing plausible answers, the more important the leadership team's ability to disagree professionally becomes.
Not personal conflict, positioning or corridor alliances, but a disciplined examination of different perspectives, where assumptions are challenged, consequences are explored and doubts can be expressed openly.
This requires trust. Not as an end in itself, but as a prerequisite for quality.
Leaders need to be able to ask:
What is this recommendation based on?
Whose interests does it serve?
What data is missing?
What is the model unable to understand?
What happens elsewhere in the organisation if we choose this solution?
And who will be accountable for the consequences?
This kind of professional disagreement is not a soft addition to business. It is part of the organisation's risk management and decision quality.
From individual intelligence to collective judgement
For decades, organisations have invested heavily in developing individual leaders: making them more strategic, more resilient, more visionary and more decisive.
Individual capabilities remain important. But strong individual leaders do not automatically create a strong leadership team.
As access to analysis and knowledge becomes more widespread, the ability to turn individual intelligence into collective intelligence becomes an increasingly important differentiator between organisations.
Connecting different perspectives. Making choices on behalf of the organisation as a whole. Standing together behind decisions and translating them consistently throughout the organisation. This is collective judgement. It is not about harmony or ensuring that everyone agrees. It is about examining disagreements thoroughly, making a decision and then acting in a coordinated way.
AI can support this process. It can provide perspectives, challenge assumptions and create a better foundation for discussion. But it cannot create commitment.
It cannot determine what the organisation should aspire to achieve. It cannot take responsibility for the necessary conflicts between legitimate competing interests. And it cannot be held accountable for what the organisation chooses to pursue or abandon.
That responsibility still rests with the leadership team.
The critical question, therefore, is not simply how organisations can integrate more AI into their work. It is how leadership teams can become better at thinking, prioritising and acting together in a world where plausible answers have become easy to produce.
AI democratises access to intelligence. It does not democratise judgement.
And yes, this article was developed with the support of AI for research and language editing – but the judgement, priorities and responsibility remain human.