Corporate Mental Health Week 1-5 July
As organisations rush to adopt artificial intelligence, much of the conversation has focused on productivity, efficiency and competitive advantage. Far less attention has been paid to what happens to people’s mental health when work becomes faster, more monitored, more uncertain and potentially never-ending.
Globally, an estimated 12 billion working days are lost every year to depression and anxiety at a cost of US$ 1 trillion per year in lost productivity. The research indicates that excessive workloads or a fast pace are among the major contributors. In South Africa, about 13% of employees live with a diagnosed mental health condition, whilst over a third of working South Africans experience excessive daily stress.
According to Prof Renata Schoeman, Head of Healthcare Leadership at Stellenbosch Business School, AI may represent one of the most significant workplace mental health challenges organisations have faced, not because machines are replacing humans overnight, but because AI is changing the pace, expectations and psychological experience of work.
“We are having extensive conversations about AI governance, ethics and cybersecurity, which are all essential. But we are largely ignoring the human consequences. AI is not only a technology issue. It is a leadership, mental health issue and psychosocial risk issue.”
More than 42% of workers say that their organisations are not evaluating the impact of AI on the organisational culture and its people. Individual adoption is also outpacing organisational readiness, with 30% of Gen Z and 31% of millennials believing their organisation is not prepared for the changes AI brings, which correlates with findings that, despite high expectations, more than 84% of companies have not redesigned jobs around AI capabilities.
Artificial intelligence promises enormous benefits. Used responsibly, it can reduce administrative burden, improve access to information and allow people to spend more time on meaningful work. But Prof Schoeman cautions that technology alone does not create healthier workplaces.
“AI will not automatically make work better. It will magnify the culture it enters. In psychologically healthy organisations, AI may free people from repetitive work and reduce unnecessary stress. In unhealthy cultures characterised by overwork, fear or poor communication, AI may simply accelerate pressure, intensify expectations and increase burnout.”
The danger, she says, is not necessarily mass unemployment. “The popular fear is that AI will replace everyone. That makes headlines, but it misses the more immediate threat. The greater risk is that AI quietly intensifies work today. Faster outputs become higher expectations, efficiency gains become pressure to do more and constant connectivity becomes normal. Within that, recovery disappears.”
“People may find themselves producing more, responding faster and being monitored more closely, while feeling increasingly exhausted, disconnected and psychologically unsafe.”
Prof Schoeman says AI anxiety is often less about technology and more about uncertainty.
“Employees do not necessarily fear the tool itself. They fear what the tool represents. They fear becoming obsolete, losing autonomy or being managed by algorithms rather than leaders. They fear that expectations will rise while support declines.”
“Uncertainty is itself a psychological burden.”
This concern is particularly relevant in South Africa, where AI disruption is unfolding against the backdrop of unemployment, economic pressure, inequality and already high levels of workplace stress.
“In South Africa, AI anxiety lands on top of existing anxiety. People are worried about finances, job security and an uncertain future. AI enters that reality. It does not enter a vacuum.”
Prof Schoeman believes organisations are moving faster than they are prepared. “Many employees are already using AI, often without clear policies, training or expectations. Leaders themselves are trying to make sense of technologies that are evolving at an extraordinary speed.”
She warns that AI-enabled surveillance and algorithmic management may create unintended consequences.
“Surveillance is not the same as support. When employees feel constantly monitored, trust erodes, psychological safety declines and autonomy diminishes. And without trust, performance ultimately suffers.”
Prof Schoeman says instead of focusing exclusively on productivity gains, organisations should ask a different question. “The question is not only whether AI improves productivity but rather whether AI improves recovery. If people are working faster, answering more emails, attending more meetings and remaining connected for longer hours, then efficiency gains may simply be disguising exhaustion.”
Burnout, she argues, should not be viewed as an individual failure. “Exhausted people do not need motivational posters. They need healthier work. And burnout is often a signal that something in the system needs to change.”
“Workplace mental health is shaped less by individual resilience and more by working conditions such as workload, autonomy, job insecurity and organisational support.”
Prof Schoeman believes psychological safety must become part of AI governance. One where a healthy AI strategy is human-centred, incorporating how AI affects workload, autonomy, fairness, trust, dignity and mental health.
“Managers will become the shock absorbers of AI transformation. They are translating uncertainty while often experiencing it themselves, ultimately determining whether AI becomes a source of empowerment or a source of fear.”
Prof Schoeman recommends the following practical steps for organisations:
- Communicate clearly about how AI will be used and how roles may evolve.
- Involve employees in decisions affecting workflows and performance expectations.
- Invest in training and upskilling rather than allowing uncertainty to flourish.
- Measure workload and recovery, not only productivity.
- Protect boundaries around after-hours work and digital availability.
- Avoid surveillance-based management approaches.
- Equip managers to discuss AI anxiety, role uncertainty and psychological safety.
- Include workload, autonomy, fairness, job insecurity and burnout within psychosocial risk assessments.
“Artificial intelligence has extraordinary potential. But if organisations use it simply to extract more output from already depleted people, we risk creating a future of work that is technologically advanced and psychologically unsustainable.”
