Looking back on my career, some of my most valuable lessons came from experiences I didn’t even realise I was learning from. Preparing a business review from scratch, working through a complex analysis, sitting in on a meeting with a senior colleague, or having my boss take apart a carefully prepared recommendation.

At the time, I was mostly concerned with getting the work done, preferably without making too many mistakes. But along the way, I learned to think through problems, question assumptions, read people, defend a point of view, and recognise when I was wrong. Much of this came from working with people more experienced than I was, and occasionally learning the hard way when I ignored their advice.

Now imagine starting your career today. You have an AI assistant that can prepare presentations, analyse data, challenge your arguments and help you rehearse difficult conversations. It’s available whenever you need it, flatters you, doesn’t get impatient when you ask the same question repeatedly, and can explain something ten different ways.

With so much knowledge and advice available, do you still need a mentor?

So, this week, let’s explore what’s driving these changes, how mentoring needs to evolve, and what we can do differently.

Why mentoring matters more in the AI era

Three developments are changing what people need from mentors and how organisations should approach developing talent.

AI is disrupting how people learn on the job.

For decades, the early part of a professional career involved plenty of routine work. Junior consultants built spreadsheets, young marketers analysed consumer data, and associates in law firms spent hours reviewing contracts.

Much of this was tedious, and I doubt many people would volunteer to do it manually again. But working through the details helped people recognise patterns, understand what could go wrong and develop an instinct for why apparently sensible recommendations sometimes failed.

AI is absorbing a growing share of this work. People can now produce sophisticated outputs much earlier in their careers without having developed the judgement to evaluate them.

As executive coach Marlo Lyons warns in a recent Harvard Business Review article:

“If the on-ramp for professional judgment disappears, so does your leadership pipeline.”

A July 2026 McKinsey article describes how one real estate firm is addressing this challenge. Although AI could produce detailed market assessments, the firm asked junior employees to visit neighbourhoods, study traffic patterns, and develop their own assessments before comparing them with AI-generated reports.

As these traditional learning opportunities diminish, managers and mentors will need to create experiences more deliberately, helping people develop judgement that they previously acquired through years of doing the work.

AI is changing what people need from human mentors.

Until recently, getting good professional advice depended considerably on whom you knew. If you were fortunate, you had a manager who invested in your development, a senior colleague who took an interest, or someone outside work you could turn to when you were stuck.

AI is changing that. Employees are using it to explore career options, prepare for interviews, practise difficult conversations and get feedback on their ideas. Someone who might never have had access to a coach can now get useful guidance at little cost.

This could make professional development far more accessible. But if people can already use AI to explore options, get explanations and rehearse conversations, what should they turn to human mentors for?

Increasingly, the answer lies in the mentor’s understanding of the individual, their ability to challenge behaviour patterns, and the experiences and opportunities they can help create. Human mentoring will need to build on what AI makes possible.

AI is changing career paths and the capabilities people need.

Roles are changing, skills are evolving, and established career ladders are becoming less predictable. The experiences that traditionally prepared people for leadership are changing too.

These challenges extend beyond younger employees. A functional leader may be figuring out how AI changes their team’s work. An experienced executive may be questioning which of their existing capabilities will remain valuable. A founder may be learning to lead an organisation where AI changes how teams are structured, and decisions are made.

There is no established playbook for many of these transitions. Mentors will need to help people navigate unfamiliar choices, experiment with new approaches and develop capabilities for roles that may not have existed a few years ago. And because younger colleagues may understand emerging technologies better than their seniors, learning will increasingly need to flow both ways.

Together, these developments call for mentoring that is more deliberate, more personalised and more open to learning in both directions.

How mentoring needs to evolve?

We’ve often thought of mentoring as a relationship in which someone with more experience shares their knowledge with someone who has less. Good mentors have always done much more, but as AI becomes a readily available source of advice, other aspects of the relationship will become more important.

Developing judgement

Consider a young manager struggling with an underperforming team member. AI can help prepare for the conversation, suggest approaches and simulate how the employee might respond.

But a mentor who knows the manager may notice something else. Perhaps this is the third time they’ve avoided a difficult conversation, or their desire to be liked is interfering with their ability to lead.

The discussion can then explore why the manager keeps finding themselves in similar situations, what they need to change and whether their behaviour improves over time.

Understanding the individual

Two people facing the same professional challenge may need different guidance. One may lack confidence, while another may be overconfident. One may need encouragement to take a risk, while another needs help recognising when to exercise restraint.

A mentor who knows someone over time can connect an immediate problem to their wider aspirations, strengths and patterns of behaviour. They can draw on previous conversations and shared experiences to help the person recognise things they may struggle to see themselves.

Creating opportunities

Some of the most valuable mentors give us opportunities to try something new, introduce us to others who broaden our perspectives, or support us as we take on responsibilities that stretch our abilities.

They might recommend us for a role, bring us into an important meeting, or vouch for us when we aren’t in the room. As AI makes advice more accessible, this part of mentoring deserves greater attention.

Learning from each other

A senior executive may have decades of experience running businesses and making difficult decisions, while a much younger colleague may understand AI tools and emerging ways of working far better.

Bringing those perspectives together benefits both. The senior leader helps the younger colleague develop judgement and understand the business context, while the younger colleague introduces new possibilities and challenges established practices.

Mentoring will increasingly involve people learning from each other, regardless of seniority.

What can we do differently?

For those of us who mentor others, this is a good opportunity to reflect on how we approach these relationships. Here are five ways you can make mentoring more meaningful and effective.

1. Build mentoring around real experiences.

Many mentoring relationships revolve around periodic catch-ups where the mentor asks how things are going and offers advice. These conversations have their place, but some of the most useful mentoring happens when there’s a real situation to work through.

Before an important meeting, ask someone what they expect might happen and how they plan to respond. Afterwards, discuss what surprised them, what worked and what they might do differently.

For younger employees, create opportunities to develop their own point of view before consulting AI. Let them analyse a problem, compare their reasoning with the AI output and discuss the differences with someone more experienced.

Some situations call for using AI from the start, while others benefit more from working through a problem independently. The challenge is to recognise the difference.

2. Make your thinking visible.

One of the most useful things an experienced leader can share is how they arrive at decisions.

When reviewing a recommendation, explain what concerns you about it. When choosing between two strategies, talk through the trade-offs. After a difficult negotiation, share what you noticed about the other side and why you changed your approach.

These informal conversations help people understand things that formal training often struggles to convey. Much of this can happen during the normal course of work, provided leaders explain their reasoning instead of simply communicating their conclusions.

3. Encourage people to build a network of mentors

We sometimes place too much emphasis on finding the right mentor, as though one person can guide us through every stage of our careers.

Different people have helped me think through different challenges. Some have been valuable sounding boards on business decisions. Others have helped me see personal blind spots, introduced me to unfamiliar industries or encouraged me to explore opportunities I hadn’t considered.

Encourage people to build a small personal advisory network, drawing on experienced leaders, peers, colleagues from other industries and younger professionals who bring fresh perspectives. AI can also play a role in exploring ideas and preparing for these conversations.

The network can change as people’s responsibilities and developmental needs evolve.

4. Use AI to improve the mentoring experience

AI can do far more than match mentors with mentees or schedule meetings.

A mentee can use it to explore options, challenge assumptions and prepare questions before meeting a mentor. Afterwards, they can reflect on the discussion and consider how to apply what they’ve learned.

Mentors can use AI too. Before a difficult developmental conversation, they can ask it to challenge their interpretation of the situation or suggest alternative approaches. Someone preparing for a negotiation can rehearse different scenarios with AI and then discuss the experience with a mentor.

5. Make developing others a meaningful part of leadership

As AI changes how employees develop their skills, managers will need to think carefully about the responsibilities they give people, the learning opportunities they create, and the feedback they provide.

Organisations can help by protecting time for mentoring, equipping leaders to have better developmental conversations and making access to good mentors less dependent on personal relationships.

We should also recognise leaders who consistently develop strong people. Look at whether the people around them are growing, taking on greater responsibilities and becoming more capable of making decisions independently, rather than counting mentoring sessions.

A final thought

AI has the potential to make good advice and guidance available to millions of people who previously had limited access to it. That’s something to welcome. But as AI becomes more capable, we’ll also need to rethink how we invest our time and attention in developing others.

For those of us who mentor others, the opportunity is to go beyond sharing our experience. We can help people develop their judgement, expose them to unfamiliar situations, open up opportunities and learn from the different perspectives they bring. And organisations need to create the time and space for these relationships to flourish.

As AI changes how we learn and work, what are you doing differently as a mentor, or looking for in those who mentor you?