Siân Harrington explores where AI coaching wins, where human coaches remain essential and what the future of hybrid coaching looks like
Jonathan Stanford never imagined that AI could coach. As head of people experience at London EV Company he’d spent years embedding coaching into leadership development, ensuring that human connection remained central to performance and growth.
So when he was introduced to an AI-driven coaching tool his reaction was instant: “I can’t see how this is going to work. I’d rather be in front of someone."
But then he watched it in action. The AI tool integrated seamlessly into workplace systems – Teams, Slack, email – offering real-time coaching insights just when employees needed them.
"I looked at the tool and thought, yes, it absolutely does work," Stanford admits. “If coaching is about raising self-awareness, helping someone to move towards their goal and making performance and behavioural change, then we’ve got to be open to thinking, surely AI could do that too."
This shift in perspective reflects a growing reality: AI is not replacing human coaching but it is changing how coaching happens.
Some coaches remain sceptical, arguing that AI lacks emotional intelligence, nuance and lived experience. Others, like Stanford, see AI as a force driving coaches to evolve.
So where exactly does AI coaching outperform human coaching? And where does the human coach remain indispensable?
Where AI coaching is proving its value
AI-powered coaching platforms are delivering results in scalability, data-driven insights and behavioural nudging.
1. AI is making coaching accessible at scale
For decades coaching was a luxury for senior leaders, expensive and time-intensive. AI is democratising access to coaching, making it an organisation-wide practice rather than a privilege for a few.
At the launch of CoachHub’s AI coach AIMY, Tom Runneson, the company’s senior behavioural scientist, highlighted figures from the International Coaching Federation:
- 109,000 accredited coaches worldwide
- 3.5 billion working adults globally
- Even if each coach worked with 20 employees per month they would reach just 2.18 million people
AI-driven coaching could bridge this gap, offering structured coaching to millions who would otherwise never have access.
"Coaching is shifting from being a corrective to a proactive tool for career growth," says CoachHub CEO Matti Niebelschütz. "AI will accelerate this shift by making coaching scalable, ensuring everyone benefits, not just high performers or struggling employees."
Joe Keilch, an executive coach for Fortune 500 companies, sees AI as particularly effective in skills development. AI coaching is great for new managers learning delegation, prioritisation and giving feedback, he says. It provides structure and accountability. But, he adds, it gets weaker when scaled: “As it gets bigger it gets further away from the individualisation that can only come from a developed relationship with another human being. And, contrary to what movies might suggest, we haven't seen humans abandoning interacting with other humans to interact with AI, or any other form of technology!”
2. AI is delivering data-driven coaching at a level humans can’t match
Measuring progress is one of traditional coaching’s biggest challenges. AI’s ability to track behaviour change in real time gives it a significant edge.
Doug Stephen, president of enterprise learning at CGS Immersive has implemented AI-driven coaching with Fortune 500 companies. He explains: "AI captures data points – frequency of skill practice, scenario completion rates, user sentiment over time – that quantify behaviour change. This data is fed back into leadership scorecards or HR dashboards, allowing organisations to calculate ROI more precisely."
At a major medical devices company AI-driven role-playing and personalised coaching improved sales teams’ success in face-to-face customer meetings. AI removes human inconsistency and ‘bad days’, says Stephen.
Arvind Rongala, CEO of corporate training company Edstellar, has seen similar impacts: "One client used AI-driven coaching to improve leadership communication. With real-time sentiment analysis and speech recommendations they saw a 25% improvement in the clarity of presentations."
"AI is already proving valuable in onboarding," says Gerti Mema, marketing manager at Equipment Finance Canada. "We use AI-powered chatbots to answer common new-hire questions, ensuring instant, consistent responses while reducing human workload." She adds that AI’s ability to identify patterns in employee behaviours offers foresight on potential issues before they escalate, though businesses are still learning how to integrate this data effectively into coaching conversations.
AI doesn’t just automate coaching but creates a measurement framework that’s been missing in traditional coaching.
3. AI has infinite knowledge – but comes with risks
AI can tap into vast knowledge sources instantly, giving coachees insights no single human coach could.
"AI can access historical data, case studies and research that a human simply cannot," says John Little, principal consultant at The Winner’s Edge Coaching.
But there’s a caveat: AI can hallucinate, generating incorrect or inappropriate responses.
Mithilesh Ramaswamy, senior engineer at Microsoft, warns: “It is not a replacement for certified human coaches who can navigate complex scenarios. AI gives a statistically probable answer but if trained on poor data it produces poor results. Basically, garbage in, garbage out."
Stephen agrees: "AI learns from historical data. If that data reflects biases AI might inadvertently perpetuate them. That’s why we conduct ‘human-in-the-loop’ audits to ensure fairness, inclusivity and relevance across diverse work cultures."
Where AI fails – and the human coach wins
For all its capabilities AI coaching has fundamental flaws. Coaches and HR leaders agree that in critical areas such as trust, emotional intelligence and transformational change AI is still miles behind human coaches.
1. AI lacks emotional intelligence and human connection
AI can simulate empathy but it cannot feel it. "It’s difficult to build empathy through a screen,” says Stanford.
Stephen reinforces this, noting that while AI can simulate realistic conversations and adapt to learner inputs, certain sensitive issues or complex interpersonal dynamics may still benefit from human intervention. For example, navigating a high-stakes conflict or dealing with acute emotional stress might require the empathy and seasoned judgment of a human coach.
Maya Gudka, an executive coach who works with London Business School, points out a deeper issue: trust. “AI is great for offering options but there can still be a gap when it comes to crystallising these into something that ‘feels’ right for the individual.”
Mike Szczesny, owner and VP at EDCO Awards & Specialties, agrees: “Human coaches guarantee safety, as well as the absence of judgment, leading people to feel free to express their ideas, problems and goals. Their nuanced understanding of tone, context and emotions ensures that coaching sessions feel deeply human and uniquely impactful.”
2. AI struggles with transformational coaching
AI excels in structured goal-setting and feedback but misses the intuitive, unscripted moments that lead to breakthroughs.
As Gudka explains: “AI might be a great place for adding information but it might not be able to create some of the identity shifts people are looking for in coaching.”
Little agrees, noting that transformational growth versus transactional advice is the biggest shortcoming of AI in the coaching space. “There are certain parts of an interaction that are unexpected, unscripted and can make all of the difference in someone having a breakthrough or getting stuck in a cycle.”
Jessica Clarke, global head of talent and L&D at Atlas Copco, sums it up: “The human coach is about shifting mindsets, holding up the mirror, looking at the body language. It’s the transformational space as opposed to the performance-based coaching of AI.”
3. AI cannot facilitate group coaching and cultural intelligence
AI coaching remains largely one-to-one. It struggles with group dynamics and cultural nuance, critical aspects of leadership coaching.
So it’s no surprise that Gudka sees group coaching as one of AI’s biggest gaps: “Group coaching allows individuals to learn, grow and connect with peers. AI is a long way from replicating that live experience."
Additionally, AI struggles with cultural intelligence. While it can be trained on language patterns it lacks deep, lived experience of different cultures.
The evidence? How AI coaching is driving self-reflection and growth
A study published in the Journal of Occupational and Organizational Psychology (JOOP) backs up AI’s impact on coaching. Led by Dr Scott Dust the study focused on users of Cloverleaf, an AI-powered coaching tool, tracking 84 professionals over two weeks to examine the relationship between automated coaching and daily reflective behaviours. The findings?
- Daily AI coaching fosters metacognitive activity, prompting users to critically assess their actions, skills and behaviours.
- Self-reflection increases the next day’s need for self-knowledge – users became more curious about their strengths and weaknesses.
- AI coaching strengthens learning orientation – individuals became more motivated to acquire new skills and take on challenges.
For HR leaders this research highlights AI coaching’s scalability, consistency and ability to drive reflective habits at scale.
The future: AI and human coaching working together
The best coaching programmes are not choosing between AI and human coaches. they are combining them.
Open innovation consultant Andrey Meshcheryakov, engagement manager at Recombinators, believes the future of AI coaching depends on understanding its role. He outlines four distinct use cases:
- AI first, coach second – AI handles routine coaching with human coaches stepping in for deeper insights.
- Coach first, AI second – A human coach leads with AI providing data-driven insights and follow-ups.
- AI complete autonomy – Low-touch coaching available 24/7, ideal for scalable skills development.
- Coach complete autonomy – Reserved for high-stakes coaching that requires human nuance.
"Not understanding these distinctions leads to subpar results," Meshcheryakov warns. "Sporadic AI use turns coaching for high performance and growth into a fancy fiasco."
He predicts that by 2025 AI-driven avatars and agent-based AI will make low-touch coaching more accessible to a wider business audience.
Stephen sees this hybrid approach as the future: “AI can handle repetitive or foundational coaching tasks, freeing human coaches to focus on more complex, higher-value areas.”
The best coaching programmes will blend AI’s efficiency with the human coach’s depth.