A friend of Christina Muller’s recently used ChatGPT to plan an outing for her daughter. The chatbot recommended a place. The description was convincing, the details were specific, and she never thought to check it. When she arrived, she found that the business had closed two years earlier. She had not searched for it on Google. She had not looked it up anywhere else. The output sounded authoritative, so she treated it as fact.
That small moment holds a much larger pattern, and it is one Christina Muller, LCSW, SHRM-SCP, spends her working life studying. A licensed workplace mental health expert and the founder of Mind Your Workplace, Muller sits at the intersection of mental health, leadership, and the future of work. She has been featured more than 30 times in national media, including Forbes, Fortune, and Bloomberg Television, and her focus is deliberately different from the conversation dominating most workplaces right now. While everyone else debates whether AI will take our jobs, Muller is asking a quieter and more immediate question: what is daily AI use doing to our minds, our confidence, and our sense of who we are at work?
The Skills You Stop Using Are the Skills Your Brain Stops Protecting
The concept Muller names first is cognitive offloading, and it is more physical than most people realize. The brain functions and strengthens itself through neural connections. When we defer to AI constantly throughout the day, at work and beyond, those connections go unused. Over time, Muller explains, they atrophy. They get weaker. The brain receives a signal that says you do not need those anymore, and a natural paring down begins.
What makes this worth paying attention to is which skills sit in the line of fire. Muller is direct about it. The abilities most at risk are creativity, discernment, and judgment, which are also the abilities she believes will never be fully replaced by AI. In other words, the fastest thing to weaken is the exact thing that keeps a person valuable.
The practical takeaway is not to abandon the technology. It is to notice which decisions you have stopped making yourself. If you cannot remember the last time you drafted something, evaluated something, or concluded without a first pass from a machine, that is the muscle worth deliberately using again this week.
Why Capable People Have Started Feeling Like Frauds
The second pattern Muller sees is one most people are not saying out loud. She describes it as imposter symptoms, a quiet reality in which people begin comparing their own work to AI output and concluding that the machine does it better. Not just faster. Better.
Muller has watched this spread past the boundaries of the job. People run their work through AI, then their emails, then their text messages, wanting to be sure everything is said perfectly and polished. The psychological cost of that habit is anxiety, and Muller says it can lead to genuine feelings of inadequacy.
Layered on top is a form of exhaustion that gets almost no attention. Because AI cannot always be trusted, the user has to check it. That constant verification, Muller says, is another layer of mental gymnastics our brains were not built to provide. It is tiring, it produces cognitive fatigue, and it can lessen creativity and affect judgment.
The research she cites reinforces the point. Muller references a Harvard study released this past spring of 1,500 employees across industries, which found diminishing returns for productivity once people are using three or more AI tools. Those employees reported what the study called brain fry: headaches, mental fog, slower decision making, and trouble concentrating. The prefrontal cortex, Muller notes, is responsible for decision making and problem solving, and it is being taxed by work it was never meant to juggle.
Automation Bias, and the Slow Devaluing of Your Own Instincts
Automation bias is the term for what happened to Muller’s friend at that shuttered venue. It is the tendency to trust the machine over your own thinking, and Muller traces part of it back to how AI was introduced. At rollout, she says, it was marketed as though AI were the smartest person in the room. That framing stuck, even as it became clear that AI can hallucinate, invent information, and invent people.
The harm compounds, because deferring becomes a habit and habits reshape self-perception. When people stop consulting their own instincts, Muller says, those instincts get devalued. It affects confidence, and it affects what makes a person unique. If everything sounds the same, she points out, we lose what is intrinsically human and interesting about the work.
That erosion is visible enough that platforms have begun reacting. Muller notes that LinkedIn recently introduced a feature allowing users to flag content they believe is AI-generated, a reversal for a platform that once built AI post-writing directly into the product. Substack has also begun screening for text it perceives as AI-generated. Both moves, in her reading, are attempts to protect real expertise, the kind that takes years to build and cannot be produced by a prompt.
“AI can’t read a room, but a human can.”
~ Christina Muller, LCSW, SHRM-SCP
When Your Role Gets Automated, Your Identity Takes the Hit
Most of us are identified by our careers. When someone asks what you do, Muller observes, you both know exactly what is meant. So when an organization tells an employee to automate a large portion of their role, the message that lands is rarely the one leadership intended. What the employee hears, Muller says, is that the company does not think they are as good as AI, and that maybe they are not as needed.
Over time, that erodes confidence and self-esteem while anxiety climbs, because the person is being told, without being told, that they do not matter. Muller connects this directly to retention data. She cites Gallup’s state of the workplace report from last year, which found people are 43% more likely to stay in a job after a couple of years if they feel they matter, and if they feel leadership cares both about them and about the work they are doing. Lose that, she says, and organizations see more burnout, more turnover, and people who feel emotionally and cognitively spent. The bottom line effects may not show up immediately, but they catch up.
Muller also names a shortage that makes all of this worse. In organizations without a democratization of AI skills and training, employees across healthcare, finance, and technology are left to automate parts of their own jobs and figure the technology out alone. That vacuum produces the exact fear she hears most often: if I do not become an expert, I will be left behind and seen as expendable.
What Leaders Get Wrong, and the Method That Takes Seconds
The most common leadership mistake, in Muller’s experience, is assuming people can simply figure it out without adequate training or any explanation of how AI complements the skills they were hired for. She is blunt about what fills the silence. When people do not know what is coming next, they fill in the gap themselves. They gossip. They spread things that are not accurate. Her counsel is to over-communicate the utility of AI and its relationship to existing skills, rather than saying less and hoping for calm.
She points to Citigroup as an example of the alternative, a case she covered in Fortune. Everyone was trained, from call center staff to managing directors, and CEO Jane Fraser framed the technology as a path to upward mobility and to more creative work rather than a replacement for the people doing it. As Muller puts it, you have to give people the tools they need to succeed, like anything else in life.
For day-to-day leadership, Muller offers a framework she calls the SEE method. You see the person and spot the effort, and you verbalize what you notice them doing well. You empathize, saying out loud that a period of disruption is genuinely hard. Then you engage in an actionable way, staying curious about what that individual actually needs in the moment, whether that is a few minutes of air, more coaching, or more support on a project. Gallup research, she notes, shows that a touch point like this even once a week has major effects on psychological safety, turnover, and whether people feel they matter. Muller tells leaders plainly that this takes seconds out of their day.
You Are Still the Pilot
The through line in Christina Muller’s work is that the answer is not resistance to the technology; it is intention with it. She compares it to exercise or eating well: something you have to be deliberate about rather than mindless about. That might mean pulling back from running your text messages through a chatbot. It might mean setting guardrails around your own use when you notice it has crept past useful.
It also means protecting connection. Muller describes the present moment as a high connectivity, low connection state, and she cites a study covered in the Atlantic finding that since 2005 we have been speaking 300 fewer words a year to one another. People race to ChatGPT instead of walking to a colleague’s desk, and in the process they stop learning what the person next to them is carrying. If your workplace does not offer those touch points, Muller says, find them elsewhere with intention, through social circles, a class, the library, or your faith. She also reminds people that support is closer than they think, through licensed therapists and through employee assistance programs, which many organizations offer as a free and confidential short-term counseling benefit of roughly three to five sessions or more.
Her closing thought is the one worth carrying into the workday. We are still the pilots of an already flying plane, Muller says, and AI can be a co-pilot, but it is not going to take over completely. You were hired because you have great skills and because someone believed you would get the job done. Christina Muller’s argument, backed by the psychology and the research she works in every day, is that remembering this is not sentimentality. It is the most practical form of self-protection available right now.
