What AI does to our mental health – and how we can work with it in a healthy way

Reading time: approx. 6–8 minutes

Studies show that 96 percent of executives expect AI to increase their teams' productivity. 77 percent of employees say their workload has increased. This sounds contradictory, but it's the current reality of AI in the workplace.

In my work with organizations, I regularly encounter this paradox: The use of AI tools is increasing, and so is burnout. What's missing is the connecting piece – an understanding of the psychological and neurobiological effects of AI on us, and what we can learn from this for healthy working practices.

The efficiency paradox: More AI, more pressure, and what that means for our mental health

An economic concept – the so-called Jevons Paradox – succinctly describes what is currently happening in many companies: When a technology becomes more efficient, its consumption increases because efficiency creates new demand. AI takes tasks off our hands – and thereby frees up capacity that is immediately filled with new expectations and tasks.

This is reflected in the numbers: According to the Upwork Research Institute, 47 percent of employees do not know how they are supposed to achieve the performance targets expected of them with new AI tools.

They are expected to deliver more – without understanding how. This isn't just a training problem. It's a structural leadership problem with direct consequences for mental health.

Neuropsychologically, acceleration has measurable consequences: Earl Miller's research (MIT) shows that our brain does not perform true multitasking, but rapid task switching – and each switch costs cognitive energy.

Sophie Leroy describes the concept of attention residue : a portion of our attention remains attached to the previous task. In an AI-accelerated work environment: permanently.

In the long term, this activates our stress system. Robert Sapolsky (Stanford) has shown that chronic unpredictability and loss of control are among the strongest stress triggers – with real physical consequences such as weakened immune system, impaired memory, and an increased risk of exhaustion.

The competence paradox: When AI makes us feel less capable

One of the least discussed consequences of AI in the workplace concerns self-efficacy – confidence in one's own abilities. Albert Bandura has shown that self-efficacy is one of the strongest predictors of motivation, performance, and mental health.

When employees experience daily how AI improves their texts and deepens their analyses – yet simultaneously fail to understand how they are supposed to achieve their set goals with AI – a dangerous vacuum arises: demands for competence without experiencing competence . This is one of the most direct paths to exhaustion and disengagement.

Added to this is what cognitive psychology calls cognitive offloading : the outsourcing of thought processes to external aids. Betsy Sparrow (Columbia University) has shown that awareness of digital availability reduces our ability to remember content.

AI goes even further: it takes over judgment, language and creative processes – areas that we perceive as deeply human.

The downward cascade: When leadership uncertainty leads to organizational exhaustion

One aspect that is often overlooked in the AI debate is that managers themselves are affected . If they don't know what AI means for their teams—and are still under pressure to deliver results—this uncertainty cascades down the hierarchy. Not out of malice, but because ambiguity spreads downwards within organizations when there is no clear direction from above.

Sapolsky's stress research makes it clear: Unpredictability is particularly stressful from a neurobiological perspective because the brain remains in a state of constant alert when it cannot assess a situation.

What helps: Amy Edmondson (Harvard) has shown that teams remain adaptive during times of change when leaders are honest about their own uncertainty. The message "We're learning this together" is not a weakness. It is one of the most effective leadership statements in transformation processes.

What healthy working with AI actually means

Working healthily with AI is not a matter of chance. It doesn't happen simply by using less AI or taking more breaks. It requires conscious decisions – on three levels.

Level 1: The individual – mental hygiene in the AI age

Actively protect cognitive breaks.

The brain needs periods of rest to process experiences and think creatively – in neuroscience, this is called the default mode network, the brain's resting mode. In a working world that is constantly freeing up more capacity through AI and simultaneously expecting more, this mode is systematically suppressed.

Alejandro Lleras-Munoz (University of Illinois) has shown that short, deliberate breaks during focused work measurably improve concentration and performance quality. The implication: Breaks are not a reward for work done; they are a prerequisite for good work.

Don't completely outsource your own thinking.

Using AI as a sparring partner is smart. Using AI as a replacement for one's own judgment is risky – not because AI is always wrong, but because we build competence through our own thinking, decision-making, and mistakes.

And according to Bandura's self-efficacy research, the experience of competence—the feeling of truly being able to do something—is one of the most important sources of psychological stability. Those who systematically outsource this experience ultimately lose confidence in themselves.

Consciously designing your own AI information diet.

Few topics are reported on as extremely as AI – either as a promise of salvation or as a doomsday scenario. Neither is psychologically helpful. Constantly consuming narratives of threat activates the stress response. Those who only consume euphoria underestimate real challenges.

A conscious selection of sources – prioritized objectively, scientifically, and with differentiation – is not a luxury, but self-protection.

Observe your own reaction to AI.

What does it trigger in you when AI writes a text that's better than yours? How does it feel when an algorithm makes a decision you would have made?

This observation – without judgment – is an important first step. Not to change the answer, but to know it in the first place.

Level 2: The Team – Psychological Safety as a Prerequisite for Healthy AI Transformation

Normalizing uncertainty about AI.

Many people in organizations don't know how to use AI effectively – but they don't talk about it because they fear weakness.

Teams need an explicit space to ask questions, make mistakes, and acknowledge uncertainty. Psychological safety is the fundamental prerequisite for this.

Consciously strengthening human connections.

The more work is mediated through AI interfaces, the more important moments of genuine human encounter become.

Neuroscientist Matthew Lieberman has shown that our brains are socially active even at rest – connection is not an add-on to work, but a biological foundation. In an increasingly digital work environment, leaders must actively cultivate this connection – not leave it to chance.

Defining together what AI will take over – and what it won't – isn't a technical decision. It's a cultural one. Teams that jointly determine what they will and won't use AI for develop a shared understanding – and retain a sense of control and purpose.

According to Deci & Ryan's self-determination theory , both are central to intrinsic motivation and psychological well-being.

Level 3: The Organization – Humanity as a Strategic Decision

Ask the crucial question.

Which tasks will AI make faster and cheaper? And which do we deliberately want to keep slower and more human-like? This question sounds simple – but it is one of the most important strategic decisions organizations are currently making.

Those who don't ask these questions let technology answer them. And technology doesn't optimize for well-being.

AI as a complement, not a replacement.

Research from the MIT Sloan Management Review shows that organizations that use AI to complement human judgment – so-called human-in-the-loop design – achieve sustainably better results than those that rely on maximum automation.

This applies not only to output quality, but also to the mental health of the people who work in these systems.

Invest in clarity, not just in tools.

The biggest source of AI-related stress in organizations is not the technology itself. It's the lack of clarity about what is expected of whom. Investing in AI without simultaneously investing in guidance, training, and psychological safety produces burnout.

Organizations that ignore this pay the price through employee turnover, quiet quitting, and declining performance quality.

Empower leaders first.

An AI transformation cannot be driven from the bottom up if there is uncertainty at the top. Leaders must first understand what AI means for their respective areas before they can provide guidance.

This is not a given: In many organizations, managers are expected to moderate change without being adequately prepared themselves.

Conclusion: Informed, conscious, humane

AI is here. It's here to stay. What we need is neither fear nor euphoria – but an informed, conscious approach.

The question "What will AI do to me?" is not a technical question. It is a question of mental health. And it deserves honest answers – at the individual, leadership, and organizational levels.

Do you want to bring this topic into your organization?

AI and mental health is no longer a niche topic – it's now appearing in every executive circle, every HR strategy meeting, every town hall, where someone asks: "And what does this actually do to our people?"

As an organizational psychologist and keynote speaker, I combine current research with what really happens in organizations – clearly, evidence-based and without buzzwords.

I speak on topics such as:

• Mental health in the AI age: What leadership needs now

• Psychological safety in times of change

• Healthy leadership – what truly strengthens teams

• Loneliness in the workplace: the underestimated health risk

Get in Touch!

💭 Reflection questions to take away:

• When was the last time you consciously did something without AI – not because you don't have it, but because you wanted to experience the process yourself? What triggered that?

• How is your team currently experiencing AI – not what they're doing with it, but how it makes them feel? Is there space to talk about this openly?

• What aspects of your organization or daily life should remain consciously human – even if AI could do it more efficiently? And have you already explicitly decided on this?

Back to the blog