In November 2022, millions of users worldwide were introduced to the chatbot ChatGPT. Earlier AI systems, such as machine learning algorithms used for candidate selection or optimising business processes, were typically introduced by the top of organisations and then communicated to employees (a top-down approach). Managers chose the systems and employees had to adapt to the new technology. But with the advent of generative AI (GenAI), that has changed.

‘What you see is that generative AI such as ChatGPT enters the organisation from the bottom up,’ Huysman explains. ‘Employees started experimenting with it at home, and before long, they were using it at work. This technology is not being imposed; it’s finding its own way into the workplace.’ This bottom-up breakthrough of GenAI has significant implications for the dynamics within organisations. While AI was previously introduced with a clear goal, such as improving efficiency or supporting specific processes, generative AI is more versatile. ‘It’s being used for all kinds of different tasks: from writing texts to brainstorming new ideas,’ says Huysman. ‘That makes it a multi-purpose tool that can promote both creativity and productivity.’

Impact on collaboration and social structures

While generative AI offers many benefits, it also comes with drawbacks and risks. According to Huysman, the change in social interactions in the workplace is one of the most notable effects of generative AI. In the past, the exchange of knowledge and ideas between colleagues was crucial for building trust and collaboration within teams. But with GenAI increasingly taking over tasks, human interactions are becoming less the norm. ‘GenAI tools such as ChatGPT make it possible to work independently without needing to consult a colleague,’ says Huysman. ‘You can simply ask such a tool to brainstorm with you or generate information that you would normally ask a colleague.’ According to Huysman, this development puts pressure on the ‘social fabric' of organisations. This informal network of trust and knowledge sharing keeps businesses running smoothly, but GenAI threatens to undermine that. Huysman: ‘In the past, if you had a knowledge-related question, you went to a colleague. Now you ask ChatGPT. But those social interactions are essential for building trust within teams. If that disappears, we will soon only have isolated employees trying to solve everything on their own.’ This is also a worrying development for new employees. ‘How will they learn from experienced colleagues if they can ask AI everything?’ Huysman wonders.

Prohibition is pointless, education is a necessity

Although some organisations are trying to keep generative AI out of the workplace, Marleen Huysman says this is an outdated approach. ‘You see that in some companies and even at universities, there’s a reflex to ban GenAI tools out of fear of the risks,’ she explains. ‘But that's not smart, because AI is already so intertwined with how humans work. Employees will continue to use it anyway, even if it is banned, and then it will happen without supervision and control.’ According to Huysman, organisations need to invest in training to ensure that employees – including managers – understand how to use GenAI effectively. ‘The problem is that many companies think they can outsource GenAI to a few experts, such as prompt engineers, and be done. But that’s not enough. Everyone in the organisation needs to learn how GenAI works and what its limitations are,’ says Huysman. ‘Managers need to understand that GenAI has become an integral part of the job, and that only with proper guidance and training can the risks remain manageable.’