Once the preserve of senior executives, corporate coaching is now a virtual, on-demand, opportunity available to all. But while, thanks to artificial intelligence, companies can offer thousands of employees this kind of job support, training experts see humans remaining an essential component of professional development.
So far, the use of AI for training has lagged behind its use in recruitment, says Alex Alonso, chief data and analytics officer at the US-based Society for Human Resource Management. “It’s an area that gets overlooked in the discussion around AI,” he says. “But the best applications live in the world of learning and development.”
Even for routine training, AI-based sessions can be more engaging than traditional formats, says Mike Mather, chief learning officer at business advisers KPMG. He sees many of the firm’s clients using AI tools to create podcasts or avatars for this kind of workplace learning.
He cites compliance training. “That’s always a challenge for people to consume,” he says. “To be able to turn that into a podcast makes the experience and effectiveness so much better than a 12-page PDF.”
AI is also changing the nature of training. While professional development once took place in scheduled sessions, it is increasingly part of the working day. By offering information and advice on-demand on mobile devices, employees can access training whenever they come across something with which they need help.
Of course, bite-sized pieces of training have been used in e-learning for many years. However, generative AI makes the experience far more interactive and conversational.
It also enables training to be designed in real time and for a workplace-specific scenario, says Matt Rosenbaum, principal human capital researcher at The Conference Board, a think-tank. “It’s accelerating the push to micro-learning modules or targeted interactions in the flow of work,” he says.
This, however, prompts a question: should companies enable generative AI to learn from interactions with employees and adapt corporate training, or should they retain some standardisation in professional learning?
“There certainly is the capacity for it to learn and change based on the queries it’s being asked,” says Rosenbaum. “But there will always be a tension, as we pursue greater personalisation, as to how we ensure everyone is hearing the same message.”
Perhaps AI’s most transformative application, however, is in making it possible to offer forms of training such as coaching to everyone, rather than just managers or senior executives.
At Bank of America, interactive simulations in which employees role-play scenarios are available in customer contact centres © Michael Nagle/Bloomberg
At Bank of America, for example, interactive, immersive simulations, in which employees role-play different scenarios, are available across the globe in the contact centres that manage the bank’s customer communications.
“Last year our teams used those tools more than 1.8mn times,” says Bernard Hampton, head of the Academy at Bank of America, the company’s onboarding and professional development organisation.
The AI Shift
Sign up to “The AI Shift newsletter”, a rigorous weekly deep-dive into how AI is reshaping the world of work, with John Burn-Murdoch and Sarah O’Connor
The training, he explains, is accessed via virtual-reality goggles or onscreen videos, and provides unlimited role playing options in simulations that mimic real-world scenarios, whether unexpected client requests or difficult conversations.
In some situations, AI coaches appear to be just as effective as their human counterparts. In one study, where participants could not tell whether they were talking to a human or a machine, no significant differences emerged in their ability to interact with the AI coaches.
Moreover, having a coach generated by AI makes it easier for employees to try things, argues Rosenbaum. “AI is not going to judge you the way a human would,” he says. “So people are going to AI first to ask the stupid questions and going to colleagues after that to have a more substantive discussion.”
Nevertheless, as AI coaches start populating the corporate landscape, companies need to ensure humans can step in if the discussion moves into sensitive areas, particularly if those conversations are being recorded.
Sensitive areas, says Alonso, include mental health topics or conversations that might put the company in legal jeopardy or risk breaches of privacy or intellectual property loss.
Technology itself can help. Systems can recognise certain words or phrases and, if deemed necessary, shut off a recording or send an alert suggesting an in-person discussion. “But these are considerations that need to be decided before unleashing this type of technology,” says Alonso.
Meanwhile, Hampton argues that humans will still be essential for the development of complex interpersonal skills. He points to leadership, conflict resolution and decision-making that requires nuanced judgment or ethical reasoning. “It’s the ability in real time to react to someone,” he says.
Mather agrees that AI experiences cannot build all professional skills. “But if the human is the professor,” he says, “the teaching assistant or tutor is going to be AI.”