Can AI build emotional intelligence for leaders, without human input
In a world where more management and leadership tasks are being replicated by AI, emotional intelligence is rapidly emerging as a leader’s superpower. But a key debate in the leadership development community is whether AI can help leaders enhance their EQ without human involvement.
Anyone working in leadership development will know that inboxes are flooded with vendors selling AI “authoring tools.” These promise to let even non-experts “design any leadership development course in minutes.” The results vary—some are impressive, others… less so.
Effective feedback course for leaders
I recently tested an AI authoring tool to design an effective feedback course for leaders. This is something we’ve been doing with humans at Threshold for 20 years. The AI’s output was very plausible. That, however, was the issue. It wasn’t insightful, thought-provoking, or original, just plausible.
It reminded me of Théâtre D’opéra Spatial, the AI-generated artwork by Jason Allen that won first place at the 2022 Colorado State Fair’s digital art competition. The win sparked fascination and controversy—some saw it as a breakthrough, while others argued it lacked emotional depth and intentionality. As AI art proliferated, its novelty wore off. Many saw it as formulaic and soulless. Over time, the hype faded, reinforcing the belief that true artistry needs human emotion, experience, and struggle. AI can mimic these qualities, but not embody them.
To be clear, AI has a role in leadership development. At Threshold, we’re using AI-generated avatars to help leaders rehearse high-stakes conversations. We call it the Flight Simulator. It works because the characters and scenarios are created by humans, drawing on two decades of experience, which also shapes the avatars’ responses.
Emotional Intelligence
So, what about emotional intelligence? Could AI help leaders improve it? First, we need to explore why EQ is becoming the essential leadership skill.
We interviewed over 40 HR and L&D thought leaders for our white paper Growing Tomorrow’s Leaders. Many agreed: emotional intelligence underpins all other interpersonal leadership skills.
Research from Harvard Business School shows that EQ explains nearly 90% of the performance gap between high- and low-performing leaders.
Emotional intelligence in adaptive leadership
McKinsey & Company found that emotionally agile leaders are 45% more effective at leading through change and uncertainty.
The International Journal of Leadership in Education reported that empathetic leaders are 40% more likely to retain top talent and 36% more likely to inspire team engagement.
According to MIT Sloan Management Review, AI can analyse emotions using text, facial recognition, and voice. But it can’t feel empathy. Human empathy is nuanced. It’s built on shared experience and genuine emotional resonance. These qualities are essential for helping leaders develop EQ. Human trainers can model real emotions in ways AI simply can’t. This offers authentic emotional interactions that help leaders learn empathy.
Understanding human emotions
Neuroscientist Lisa Fieldman Barret, author of How Emotions are Made argues that the expression of emotions, in speech and facial expression, does not follow the pre- determined patterns on which AI programmes build their assumptions. On the other hand, actual humans have an extraordinary ability to read emotions in others, even when they are not obviously following a commonly understood lexicon.
AI, no matter how advanced, operates on algorithms and data patterns. It can simulate empathetic responses but it can’t genuinely experience or understand human emotions.
AI is not able to grasp the full complexity of human emotions. A study by Harvard Business Review (HBR) found that emotionally intelligent feedback from human mentors led to a 25% greater improvement in leadership effectiveness than AI-driven feedback alone.
Jonathan Berry is UK Shadow Minister for AI and Intellectual Property and a member of the House of Lords. He argues that large language models can simulate humans reasonably well. They sound perfectly plausible. But we should never forget that they are not human and ultimately cannot do the jobs humans do.
“It’s a bit like eating wax fruit and expecting to be nourished.”
It’s worth reminding ourselves that these are the words of one of the UK’s leading advocates of the technology, who served as the government minister for AI in the previous administration. He reminds us that AI can replicate a ‘difficult conversation course’ very plausibly. However, to participants, it will always feel a little like ‘wax fruit,’ even if they are not consciously aware of it. There is no substitute for the human mind that has evolved to navigate the real-world social landscape of humans over millennia. The solution is the intelligent combination of AI technology with genuine human input.
