Why AI won't be the burnout cure we've been waiting for

Aug 16, 2023

Since ChatGPT launched to the public in November 2022, workers have flocked to the platform; social media feeds have filled with tips on how to incorporate AI tools into workflows; and some early research has already shown generative AI leads to productivity spikes and increased job satisfaction.

Findings from Microsoft's 2023 Work Trends Report echo this excitement, suggesting these tools will lighten heavy workloads and potentially help employees deal with burnout. And many workers seem to agree. In a survey of more than 6,000 global respondents from enterprise automation-software company UiPath, 58% of respondents said they believe automation can address burnout and improve job fulfilment.

Burnout is still a widespread workplace issue, according to experts. Many find the prospect that these emerging tools could automate mundane, time-consuming tasks a promising development – and the news is a bright spot amid worker concerns around the rise of generative AI.

Yet they also say there's a caveat. As workers use these AI tools to increase productivity, it's not clear whether lightened workloads and less drudgery will enable workers to take a breath, or simply create more space to fill up with new tasks.

Many experts agree that busywork – such as answering emails and attending meetings – is a contributor to burnout. "People have a lot more repetitive work to do now," says Ed Challis, head of AI strategy at UiPath.

Challis believes AI tools have the potential to cut down on the busywork. "It really will be like every single person has a personal assistant," he says. "So, a lot of the repetitive work we do at work will disappear."

And, while it's too early to know the precise impact AI tools will have at work, results from a working paper from the National Bureau of Economic Research, a Boston-based non-profit organisation, shows the productivity-boosting potential of the technology.

The authors, from Stanford University and the Massachusetts Institute of Technology (MIT), recorded a 14% increase in productivity of customer service representatives working with an AI chat assistant. The AI assistant narrowed the gap between the top and worst performers, and also accelerated the training of new employees. One customer service agent with only two months experience, helped by the AI assistant, performed just as well as an agent with six months of experience working alone, according to the study.

Companies will have to choose how to use the increased productivity AI might bring by either focusing on the bottom line, or prioritizing workers' health (Credit: Getty Images)

However, the prospects for long-term outcomes – like mitigation of burnout – aren't as straightforward. Danielle Lee, one of the paper's authors, agrees that AI tools can aid productivity and cut down busy work, but he can't predict the precise way those factors will affect workers.

"If most of my time was spent doing something drudgerous, and now I’m doing more interesting things, that might be good from a burnout perspective," says Lee, an associate professor of technological innovation, entrepreneurship, and strategic management at MIT Sloan School of Management. "But it's also unclear whether being more productive means you actually get a break. It could just be the case that these workers are able to handle more in a given day."

Lee also suggests that, if employees become more productive, employers may decide they need fewer of them – adding even more heft to individual workloads. "It's very easy for companies to keep raising the bar," says Lee. Burnout remains an organisational problem, she explains, and while technology can impact it, the way it's ultimately addressed is up to the company. "You can use AI to ease burnout, or you can use it to make it much, much worse," says Lee.

Companies have a choice, she says: they can either share the benefits of the increased productivity, by, for example, cutting down work hours; or they can focus entirely on the bottom line. "This is an accelerator button that we can push," says Lee, "and companies can just keep pushing it."

Yehuda Baruch, professor of Management at University of Southampton Business School, UK, agrees these AI tools aren't a silver bullet for solving burnout. He predicts AI will remove the need for many of the jobs that humans currently do today, and that those left working will be mostly focused on creative and innovative tasks.

But Baruch points to another issue: experts say that burnout isn't just caused by overwork or repetitive tasks. Findings from a 2021 study by Baruch and academics in China and Hong Kong showed hospitality workers were more likely to get burnt out when they became anxious over AI potentially taking their jobs.

"Burnout can come in many different forms," says Baruch. "When you see all the knowledge you’ve gained, and skills you’ve developed, over years, being done by a machine, it will be frustrating." This burnout won't be caused by having too much to do; instead, Baruch explains worrying you are no longer needed could take a similar toll.

In their book The Burnout Challenge, psychologists Christina Maslach and Michael Leiter describe burnout as not an "outcome of a single factor, but a complex mix of exhaustion, cynicism, and inefficacy". In the six distinct causes of burnout they identify, overwork is merely one.

The UiPath research also shows workers reporting various root causes for burnout. Although many respondents pointed to common overwork factors such as "working beyond scheduled hours" (40%) and "too many meetings or calls" (25%) as at least one source of their burnout, they also cited workplace culture, such as "pressure from leadership" (39%), and job insecurity.

Ultimately, the experts believe employers will have a significant role in how AI can improve workers’ mental health – for instance, whether they load up workers with more tasks and cut staff, or whether they enable workers to reclaim their time instead. And that uncertainty – regardless of how much tedious work employees can expect to offload – can contribute to burnout in and of itself.

As Baruch says: "AI could be a blessing, but it could also be a curse."