Coaching and AI: Bridging the gap between artificial and authentic
By Jo Faragher on 25 July 2024
Imagine a scenario where there is an emerging leader in your organisation. Perhaps they’re not the traditional coaching candidate or budgets do not extend to offering them a full package of support. But they could really benefit from connecting with someone on their career aspirations and the personal and professional barriers they need to overcome to meet their goals.
Whenever they need that support, they could ask an ‘AI coach’ for answers and receive them 24/7, without the cost and administration of bringing in a human expert. In theory, an AI coach seems to solve a lot of issues with traditional coaching – it can be rolled out more widely at low-cost, it can give immediate responses, and handle a lot of the basic tasks associated with coaching, such as scheduling, sending reminders and summarising the conversation.
However, most organisations are rightly apprehensive about the use of AI in something that is such a ‘human’ practice. After all, good coaches help coachees ask themselves the right questions through empathy and trust, which AI tools do not have. Human coaches don’t tell their coachees what to do; they help them find out those answers for themselves. “A human coach knows you and will challenge you in a way that goes deeper,” argues Rebecca Hourston, global head of coaching at Talking Talent. “They can bring in history and context – and most of all, empathy, accountability and that indefinable inspiration that comes from authentic human-to-human support and connection.” Pitted against each other, there seems to be a clash between the artificial (the AI) and the authentic (the human coach).
Working together
But could there be a role for AI to sit alongside a human coach? Coaching and technology providers are increasingly customising their tools to offer a more personalised experience and embrace the best of both worlds. They argue that having an “always-on” coach available via a digital platform means a coachee can access spontaneous support when needed, for example if someone has just had a difficult conversation or is facing a dilemma and needs quick suggestions. “An AI coach can nudge the coachee, prompt them, get them ready for their next live meeting so they get the most out of it,” explains Hannah Priest, head of product, research and innovation at Talking Talent. “The data the AI coach collects is hugely valuable and helps the live coach really get to where they need to, and for clients they have a range of data showing them the challenges their people are facing,”
But how ‘intelligent’ is the AI coach, and are there risks associated with asking a robot for support with an important workplace issue? If the data that goes into the algorithm is sophisticated enough, the responses an AI coach is able to offer will be all the more relevant and useful to the person on the other end, adds Priest. “In building our AI coach, we have emotionally intelligent coaches training the AI. And it’s enclosed, so the responses and actions are only based on our own coaching practices and materials.” Alongside a human coach, an AI tool can amplify the impact of the human coach and make the outcomes more sustainable, she says. For example, habit creation can be a challenge within the coaching relationship, as work deadlines and life can get in the way. An AI coach can nudge someone, check in on their progress and celebrate it.
Psychological safety
There are further advantages of blending live coaching with AI. Younger employees or those who are neurodivergent may prefer the psychological safety of seeking guidance from a computer rather than a human, for example (although there will also be employees who feel less safe with an AI, so this needs to be considered). “For some people working with an algorithm will feel more comfortable, and for people who have had less of a voice or who struggle to organise their thoughts it may be particularly helpful,” adds Hourston. “It can help people avoid that dreaded blank page, giving them somewhere to start. You could even use it to inject lightness into a task or goal, such as asking it the ‘top 10 worst things a manager could do in this situation’, as a starter to push off against and help you think through the best, most productive courses of action instead.”
Andy Pardoe, a professor of artificial intelligence and founder of tech company Wisdom Works, believes AI could revolutionise company coaching in terms of boosting its accessibility. “Through AI-powered platforms and AI powered digital coaches (AI agents), organisations are able to scale coaching initiatives to reach a larger number of employees, leading to highly personalised guidance and support at a fraction of the usual cost, “ he says. But he concedes that this needs a “measured approach” where each company identifies its specific coaching needs and chooses suitable tools based on this. He also recommends providing in-depth training on how employees can use such resources in a responsible way. “From my perspective, regular monitoring and feedback mechanisms are essential to ensuring the continued reliability and efficacy of AI-driven coaching programmes,” he adds.
AI can also help with accountability between a coach and their coachee, sending out automatic prompts or questions that keep the employee on track with their goals, or scheduling catch-ups so they don’t get missed due to everyone’s packed schedule. Innovative applications of AI in coaching could include summarising feedback and identifying recurring patterns. “One of the dangers of this is that it could make the knowing-doing gap even bigger, however,” adds Hourston. “AI is great at giving us more, but often what we most need is less. AI can be very prescriptive (here are the 20 things you should do to….), which masquerades as coaching but really is just offering advice. A great human coach will avoid advice altogether, will ask what is the one thing a coachee most needs to do, and will help them move into tangible action and clear the roadblocks to that.” Of course, AI is always learning and developing its ‘intelligence’, so conversations will become more carefully curated over time.
This can also be a valuable selling point when trying to get buy-in with senior leadership teams. Alan Lerner, professor in digital business at the Open Institute of Technology, says that one of the big selling points is being able to offer data-driven insights. “Analysing individual data like personality traits, habits, or past experiences will help coaches tailor their approach based on individual needs, avoiding a ‘one-size fits all’ approach. Saving time by automating repetitive tasks such as scheduling appointments or sending email reminders means human coaches can spend more time helping individuals, he adds.
That said, any organisation considering the use of AI in coaching, ethical considerations should be a priority. It’s well documented that algorithms can be biased thanks to the data that populates them and the fact that development teams, historically, have been less than diverse. “It’s all about the way you train your AI,” adds Priest. “We use coaches with a diverse range of views, and ensure our testing process has diversity built in, and is part of how we iterate. The more the platform learns from the coaches, the more diverse the range of perspectives it builds.”
Augment and amplify
Using AI to amplify the impact of human coaching, rather than seeing them as mutually exclusive, could be a positive way forward. It’s about maintaining the core human elements of coaching such as building empathy, drawing on lived experience and holding someone to account, while calling on AI tools to offer customised suggestions and carry out less high-stakes elements of the process. For example, an AI can be used in the matching process to ensure that coaches and coachees are well matched based on their needs and experiences, and can even set up the relationship through meeting requests and the like. Pardoe says that alongside ensuring there are regular audits and refinements of the platforms used, human involvement must remain a valuable component. “I would argue that a sympathetic human touch is important for establishing trust, rapport and engagement between coach and coachee. Leveraging a hybrid approach with human coaches augmented with AI coaches in support might be the best solution longer-term,” he advises.
Whatever the ideal blend of human and AI coach is, organisations looking to embrace this in their workforce development need to move at pace. As we’ve seen with the latest iterations of generative AI (ChatGPT) tools, it can perform ever more sophisticated tasks and we need to work with, rather than against, this tide.
View more d&i Leaders articles
d&i Leaders is a global community of senior diversity, inclusion and HR focused professionals, looking to collaborate, network and accelerate their workplace inclusion strategy.