mentoring and supervision for human coaches will still be viable at least in the next 5 years but I am concerned if these 2 aspects of reflective work of a coach will also be replaced by AI. In the mentoring space, I note that the use of tools such as Raenotes can pick up coaching competencies in an audio recording with about 60% accuracy. ChatGPT, Claude, etc. can engage in intellectual discussions on issues faced by a coach and provide almost similar support as a human mentor and supervisor.

Mental health awareness and support both are increasing rapidly. However, mental health specialities are not as discreetly established as those of physical health.
Boundaries between therapy and coaching are often faded by clients as well as coaches. Since, coaching does not have practice specific educational, licensing or certicational requirements, it risks the speciality of the practice.
I feel coaching industry will benefit from regulations.

the coaching profession will be the number one human services intervention that people and institutions seek relative to varied personal and professional challenges. As the profession continues to develop and evolve, the evidence, results, and effectiveness of coaching set a renewed precedent about the industry. Making coaching a viable human services intervention option.

coaching practices are evolving at different speeds across regions. In the north America and Europe, the field is more mature, especially within large corporations and multinationals. In the Middle East, coaching is still developing, with a growing interest in accredited training programs. Government organizations, in particular, are investing in upskilling their executives and HR professionals to become manager-coaches and internal coaches, often through large-scale programs.

as soon as a larger player in tech gives ai coaching bots sensors (to see and hear) we will see a dramatic shift in capabilities and adoption. A software/ hardware combination in your device camera linked to the “office suite” you use makes the feature standard and overcomes adoption decisions. Current limitations are due to trying ti force a text interface to do something that was never done by text.