Why AI can be a powerful first mirror
Over the past few months, several conversations with clients have pushed me to reflect more deeply on how artificial intelligence is changing the way professionals think about their careers.
Many people I meet today have already spoken to an AI chatbot before speaking to a coach.
They have used it to make sense of a difficult situation at work, prepare for a salary discussion, analyse a conflict with a colleague, explore a career change, understand a loss of motivation, or simply put words on a growing sense of fatigue, doubt or misalignment.
Some have even compared the answers generated by AI with what emerged during our coaching sessions. Not in a provocative way, but with genuine curiosity. What does AI do well? Where does it fall short? What changes when a real person is present in the conversation?
These exchanges have been honest, sometimes vulnerable, and very revealing. They show something important: professionals are not rejecting human support. They are trying to understand where technology helps, where it reassures, and where it cannot go far enough.
This is not a theoretical question. It touches the heart of career development in a changing labour market. In Switzerland and across Europe, professionals are under pressure to adapt faster, remain employable, make better choices, and navigate increasingly complex organisational realities.
AI offers immediate help. But immediacy is not the same as depth.
The attraction of AI is easy to understand
It is available at any time. It does not judge. It structures thoughts quickly. It can generate a salary script, a list of transferable skills, a re-skilling plan, a difficult-conversation framework, or a first version of a CV in a few seconds.
For many people, this is useful. Sometimes very useful.
Recent research from The Conference Board found that AI can provide up to 90% of day-to-day coaching functions, especially around structure, role-play, goal-setting, feedback and next steps. The same research reported that 96% of surveyed workers felt AI responses were tailored to their goals or context, while 89% said the session generated specific and useful next steps.
This confirms what I see in practice. A person who feels stuck at 11 p.m. can ask a question without waiting for an appointment. Someone who is afraid to speak openly about burnout, frustration, lack of recognition or professional doubt can test their thoughts privately. Someone preparing for an interview can rehearse, refine and organise their arguments.
In that sense, AI can act as a first mirror. It helps people move from confusion to structure. It gives language to thoughts that were still scattered. It can lower the entry barrier to reflection.
AI as a formulation tool
I see this regularly. Clients arrive better prepared because they have already explored certain questions with AI. They have identified possible options. They have drafted arguments. They have clarified part of the problem.
One client illustrated this very clearly.
Before each session, he would ask Claude to help him reflect on the topic we were going to discuss. Sometimes he even asked: “What do you think Leslie, my coach, would say about this?”
For him, this was not about replacing the coaching conversation. It was a way to arrive better prepared. He felt he was bringing more material into the session: more questions, more hypotheses, more angles to test.
It also allowed us to compare the AI-generated perspective with the human coaching perspective, not as a competition, but as a constructive confrontation.
This became valuable because the real work happened in the debrief. Together, we could examine what was useful, what remained too generic, what was missing, and what needed to be challenged.
Conclusion
AI helped him formulate.
The coaching session helped him interpret, prioritise and decide.
This is perhaps one of the most promising uses of AI in career coaching: not as a substitute for human guidance, but as a preparation tool that makes the human conversation richer, more precise and more useful.
That is progress.
But it is not the whole process.
And this is where the real question begins: when does AI stop being sufficiently helpful, and when does human discernment become essential?
In the second part of this series, I will explore why the real risk is not using AI, but using it without discernment, and why better answers are not enough if we are not asking the right questions.
