A chance conversation with a US academic recently turned to how students are using AI in assignments. No surprise, it is already becoming a significant issue.

Some students use AI carefully as a thinking partner or research assistant. Others simply copy and paste entire outputs into submissions, occasionally including ChatGPT’s accidental footer: “Would you like me to take this to A-grade level?”

It made me wonder:

How are leaders using AI right now?

At Mantle, we are already using AI to augment learning design, accelerate content development, and improve productivity. Most organisations are doing the same.

But a much bigger question is emerging:

What happens when AI starts influencing organisational judgment itself?

The decision-making problem AI is walking into

Research from Deloitte found that 60% of executives now regularly use AI to support decisions, yet only 5% believe they are leading in AI-enabled decision-making capability. That gap matters.

Because AI is arriving at precisely the moment many leaders are already cognitively overloaded.

The same research found that 85% of leaders regret or question major decisions they have made, while 72% say the sheer volume of data, and lack of trust in it, has delayed or prevented decision-making altogether.

Under those conditions, humans become highly vulnerable to automation bias: the tendency to over-trust automated recommendations, particularly under pressure.

Busy leaders. Compressed timelines. Too much information. Decision fatigue.

Under pressure, AI can become cognitively seductive…like GPS for the brain.

The risk is amplified because many organisations did not have strong decision-making disciplines before AI arrived. Decision Design estimates businesses may lose up to 40% of profits through poor decisions, while McKinsey reports many executives believe their organisations are too slow and inconsistent in making major decisions. Deloitte’s 2026 Human Capital Trends research also found that 57% of organisations operate at low decision-making maturity before AI enters the equation at all.

Now many of those same organisations are accelerating decisions using systems that create the appearance of intelligence.

That should concern leaders more than it currently does.

AI does not replace weak judgment. It amplifies it.

One of the most important findings in recent AI research came from Harvard Business School. Researchers found AI improved the performance of high-performing entrepreneurs by 10–15%, while reducing the performance of lower performers by 8%.

AI acted less like a universal capability uplift and more like a cognitive amplifier.

Good thinking became better.

Weak thinking became faster.

That distinction matters enormously.

One of the biggest misconceptions in the AI conversation is the assumption that access to intelligence automatically improves judgment.

It does not.

AI can generate options, summaries, forecasts, recommendations, and even strategic analysis. But like any amplifier, it magnifies what already exists underneath; strong thinking, weak thinking, bias, discipline, shallow analysis, and ethical blind spots.

AI can support analysis.

It cannot carry accountability for consequences.

Humans still do.

The real danger is not that AI starts thinking for us. It is that we slowly stop thinking rigorously ourselves.

A useful question for leadership teams may now be:

Where in our organisation are people starting to outsource thinking rather than augment it?

Decision fitness may become the defining leadership capability

The organisations that thrive in the AI era may not be those with the most advanced tools. They may be the ones with the strongest judgment disciplines.

What matters increasingly is decision fitness; the ability to maintain high-quality judgment under conditions of speed, ambiguity, complexity, and AI augmentation.

That means leaders capable of critical thinking and ethical discernment, but also intellectual humility; the ability to question assumptions, tolerate ambiguity, and recognise when human override matters most.

Importantly, this is not about slowing organisations down.

It is about understanding which decisions can move fast, which require challenge and debate, and where AI should advise rather than decide.

Data can reduce uncertainty.

But it cannot eliminate ambiguity.

That remains profoundly human territory.

The philosophical risk leaders are underestimating

One of the most important AI articles of 2025 was MIT Sloan Management Review’s Philosophy Eats AI. It asked a deceptively simple question:

Are leaders making AI-enabled decisions that reflect their organisation’s values, or defaulting to the philosophies embedded inside the models themselves?

Because every AI system contains assumptions about what matters, what success looks like, what risk means, and what trade-offs are acceptable.

If leaders never interrogate those assumptions, they are not just outsourcing analysis.

They are outsourcing worldview.

Final thought

Gartner projects that by 2027, half of all business decisions will be augmented or automated by AI agents.

The next few years may become the critical window in which organisational judgment either strengthens dramatically, or slowly weakens through disuse.

This is not just a technology story.

It is a leadership capability story.

Or more bluntly:

This is a leadership development emergency disguised as a technology transformation.

The real risk is not that AI becomes more intelligent than leaders.

It is that leaders become less intellectually disciplined because AI became cognitively convenient.

Your AI does not have judgment.

You do.

Like to know more about how to use practical neuroscience to drive more impact?

Register for one of our public leadership programmes

Connect with us on LinkedIn, or sign up to our e-newsletter