When your boss asks ChatGPT, who can you trust?
Artificial intelligence is beginning to influence decisions in executive offices. But as managers increasingly rely on AI to make decisions, they risk undermining trust between leaders and employees. That is the warning from Professor Alf Rehn at the University of Southern Denmark.
Nowhere in the world do people trust one another more than in Denmark. According to international data from the World Values Survey, approximately 74 per cent of Danes believe that most people can be trusted.
But artificial intelligence could pose a challenge to Denmark's culture of trust. What happens to trust between people when we can no longer be certain who is actually behind the words and decisions we encounter?
Take your boss, for example. When faced with a difficult decision about a new strategy, a restructuring or perhaps an employee's future, they might previously have discussed the matter with colleagues or fellow managers. Today, there is a new adviser who is always available. ChatGPT.
And this is no longer just a hypothetical scenario. Professor Alf Rehn from the University of Southern Denmark says that several senior executives he works with are already using generative AI to support their decision-making.
"When I've built up enough trust with a CEO or senior executive, they sometimes admit that some of their decisions have been made after consulting generative AI," says Alf Rehn.
He has coined the term vibe management to describe a new approach to leadership in which managers increasingly use generative AI to inform their decisions.
But what happens when the technology moves beyond offering advice and effectively starts making the decisions?
"If managers are making their decisions through generative AI, why do we even need them?" asks Alf Rehn.
However, problems arise when employees begin to question whether their managers are genuinely responsible for the decisions they make.
"Vibe management is also about suspicion. What happens when employees begin to suspect that management doesn't really know what it's doing, but is simply feeding questions into a machine and repeating whatever answers come back?" says Alf Rehn. And that mistrust can work both ways.
Where managers could previously assume that employees had carried out their own work, it has now become much easier to delegate tasks to artificial intelligence.
This could create a situation in which managers question their employees' contributions, while employees question their managers' judgement.
"We often forget how important trust is in society, particularly in a country like Denmark, where trust between people is exceptionally high. We even leave prams outside cafés without worrying that someone might take our children," says Alf Rehn.
He has experienced this shift himself in his work with university students.
"I never used to doubt my students. But now, when I receive an assignment, I use a tool to check whether there might be an issue with AI-generated text. It's actually rather sad, because it illustrates how trust is beginning to be challenged," he says.
According to the professor, there is still far too little discussion in Denmark about what happens to our culture of trust when AI becomes a natural first choice in both professional life and education.
Established in late 2024, the centre investigates how data, algorithms and artificial intelligence are transforming organisations, leadership and society.
"We're not particularly concerned with exactly how an AI system is programmed. We're interested in what happens when these incredibly powerful technologies become part of human organisations. How do they change leadership, and what ethical dilemmas do they create?" Alf Rehn explains.
He has also previously floated a provocative thought experiment about replacing a substantial proportion of middle managers with psychologists. The idea is that AI could take over many administrative and coordinating tasks, while employees' human needs require skills that traditional managers do not necessarily possess.
Three and a half years later, he acknowledges that change has been slower than he anticipated.
"It's remarkable how slowly things have changed. Given how powerful AI has become, fewer things have changed within organisations than I would have expected three years ago," he says.
Nevertheless, he still expects significant changes in the coming years. During a recent visit to Meta in Ireland, he encountered open discussions about efficiency improvements, reducing management layers and creating flatter organisational structures.
"Middle managers need to seriously consider what they actually contribute and what value they bring to an organisation. The old days of simply drinking coffee and telling employees what was said at the latest meeting are coming to an end," says Alf Rehn.
Yet while AI may take over more administrative and coordinating responsibilities, organisations also risk losing some of the valuable functions that middle managers have traditionally performed.
Perhaps, then, the most important question about the future of leadership is not simply how many tasks artificial intelligence can take over, but what happens to trust between people when it does.
But artificial intelligence could pose a challenge to Denmark's culture of trust. What happens to trust between people when we can no longer be certain who is actually behind the words and decisions we encounter?
Take your boss, for example. When faced with a difficult decision about a new strategy, a restructuring or perhaps an employee's future, they might previously have discussed the matter with colleagues or fellow managers. Today, there is a new adviser who is always available. ChatGPT.
And this is no longer just a hypothetical scenario. Professor Alf Rehn from the University of Southern Denmark says that several senior executives he works with are already using generative AI to support their decision-making.
"When I've built up enough trust with a CEO or senior executive, they sometimes admit that some of their decisions have been made after consulting generative AI," says Alf Rehn.
He has coined the term vibe management to describe a new approach to leadership in which managers increasingly use generative AI to inform their decisions.
But what happens when the technology moves beyond offering advice and effectively starts making the decisions?
"If managers are making their decisions through generative AI, why do we even need them?" asks Alf Rehn.
When trust begins to erode
Alf Rehn emphasises that AI also has the potential to improve leadership. For example, the technology may help identify biases and blind spots that human managers fail to recognise in themselves.However, problems arise when employees begin to question whether their managers are genuinely responsible for the decisions they make.
"Vibe management is also about suspicion. What happens when employees begin to suspect that management doesn't really know what it's doing, but is simply feeding questions into a machine and repeating whatever answers come back?" says Alf Rehn. And that mistrust can work both ways.
Where managers could previously assume that employees had carried out their own work, it has now become much easier to delegate tasks to artificial intelligence.
This could create a situation in which managers question their employees' contributions, while employees question their managers' judgement.
"We often forget how important trust is in society, particularly in a country like Denmark, where trust between people is exceptionally high. We even leave prams outside cafés without worrying that someone might take our children," says Alf Rehn.
He has experienced this shift himself in his work with university students.
"I never used to doubt my students. But now, when I receive an assignment, I use a tool to check whether there might be an issue with AI-generated text. It's actually rather sad, because it illustrates how trust is beginning to be challenged," he says.
According to the professor, there is still far too little discussion in Denmark about what happens to our culture of trust when AI becomes a natural first choice in both professional life and education.
Research into AI and human leadership
These are precisely the kinds of questions Alf Rehn and his colleagues are exploring at CODES, a research centre at the Faculty of Engineering at the University of Southern Denmark.Established in late 2024, the centre investigates how data, algorithms and artificial intelligence are transforming organisations, leadership and society.
"We're not particularly concerned with exactly how an AI system is programmed. We're interested in what happens when these incredibly powerful technologies become part of human organisations. How do they change leadership, and what ethical dilemmas do they create?" Alf Rehn explains.
Sack the middle managers and hire psychologists
Back in 2023, Alf Rehn predicted in an interview with the University of Southern Denmark that artificial intelligence could make much of the work performed by middle managers redundant.He has also previously floated a provocative thought experiment about replacing a substantial proportion of middle managers with psychologists. The idea is that AI could take over many administrative and coordinating tasks, while employees' human needs require skills that traditional managers do not necessarily possess.
Three and a half years later, he acknowledges that change has been slower than he anticipated.
"It's remarkable how slowly things have changed. Given how powerful AI has become, fewer things have changed within organisations than I would have expected three years ago," he says.
Nevertheless, he still expects significant changes in the coming years. During a recent visit to Meta in Ireland, he encountered open discussions about efficiency improvements, reducing management layers and creating flatter organisational structures.
"Middle managers need to seriously consider what they actually contribute and what value they bring to an organisation. The old days of simply drinking coffee and telling employees what was said at the latest meeting are coming to an end," says Alf Rehn.
Yet while AI may take over more administrative and coordinating responsibilities, organisations also risk losing some of the valuable functions that middle managers have traditionally performed.
Perhaps, then, the most important question about the future of leadership is not simply how many tasks artificial intelligence can take over, but what happens to trust between people when it does.