Before blaming AI, look at how your business makes decisions

We worry that artificial intelligence will stop people thinking for themselves. It is a fair concern. But business owners should also ask whether their workplaces encourage independent thinking in the first place.

Most of us have attended the meeting where everyone contributes, everyone agrees and nobody makes a decision.

We promise to 'circle back'. Someone writes a summary. Another meeting goes into the diary.

Three weeks later, the problem is still there.

Adding AI to that process might produce a faster summary. It will not automatically produce better judgement.

For Lancashire businesses considering how to use AI, the starting point should be the quality of the decisions it will support.

If the owner announces the preferred answer before asking for opinions, agreement tells you very little.

The person closest to the customer might see a problem that the management team has missed. A colleague might know why the proposed solution will fail, but feel uncomfortable challenging someone more senior.

Before backing a recommendation, ask: “What evidence would make us change our minds?”

Then give people room to answer. Challenge the proposal without making it a test of loyalty to the person proposing it.

AI can help examine assumptions, compare alternatives and identify questions worth investigating. It can also confidently produce errors or build a convincing argument around a weak starting point.

Ask it to explain why your preferred plan is right and you may receive exactly the reassurance you wanted.

A more useful approach is to ask:

• Which assumptions does this proposal depend on?
• What evidence is missing?
• What is the strongest argument against it?
• What alternatives have we overlooked?
• What would make this fail?

Treat the answers as material to investigate. Check factual claims and test the reasoning against your business, customers and commercial circumstances.

At your next management meeting, try recording seven things:

• The problem being addressed.
• The evidence available.
• The options considered.
• The decision reached.
• The person responsible.
• The deadline.
• How success will be assessed.

If a decision cannot yet be made, record what information is missing, who will obtain it and when you will return to the question.

That gives the next conversation a purpose.

Sometimes the obstacle is an approval process that prevents capable people from acting.

If someone is responsible for an outcome, be clear about which decisions they can make, what budget they control and when they need to escalate.

A faster tool cannot resolve unclear authority.

I explore the tension between AI, corporate agreement and independent thought in AI Is Killing Our Humanity. Have You Been to a Management Meeting?

Business decisions affect money, jobs and livelihoods. They deserve evidence, constructive challenge and someone prepared to take responsibility.

Use AI to help people think more carefully. Keep people accountable for what happens next.