AI psychosis is the new leadership blind spot
Last year, a psychiatrist at the University of California, San Francisco, reported that he had hospitalized 12 people in a single year after they had “lost touch with reality because of AI .” He called it “AI psychosis.” It isn’t an official diagnosis, and cases that severe remain rare. But the mechanism behind it deserves every leader’s attention. Because what loosened those patients’ grip on…
In a recent report, a psychiatrist from the University of California, San Francisco, announced the emergence of "AI psychosis." This term, although not officially recognized, refers to a state where individuals lose touch with reality due to overexposure to AI. While severe cases remain rare, the underlying mechanism poses significant concerns for leaders worldwide.
As AI continues to advance, executives are increasingly relying on its advice, with 74% of executives expressing more confidence in AI's guidance than that of their colleagues or friends. Furthermore, 44% of senior leaders would defer to a chatbot's reasoning over their own insights, a troubling trend that warrants attention.
The primary concern surrounding AI's influence on leadership is its potential to foster excessive trust. Researchers have identified three key symptoms that indicate this growing dependency. Firstly, leaders are increasingly bypassing critical checks on AI-generated content. As people place more faith in AI, they tend to apply less scrutiny to its outputs, resulting in AI-drafted board updates, emails, and strategy papers receiving only a cursory glance.
This phenomenon, dubbed "workslop," leads to polished but often superficial content. Moreover, employees who receive AI-drafted emails perceive their senders as less trustworthy and intelligent, undermining the intended benefits of AI-enhanced communication.
Secondly, there is a troubling trend of mandating AI usage before establishing proper governance. In 2025, numerous CEOs declared AI a baseline expectation, with some companies requiring managers to prove AI's incapability before hiring human alternatives. However, a recent survey revealed that 78% of executives lack confidence in passing an independent audit of their AI governance within 90 days.
While executive teams publicly support AI initiatives, few possess the necessary mechanisms to effectively monitor, own, and control these systems, leaving vulnerabilities unaddressed.
The most alarming consequence of this AI-driven overconfidence is the tendency to trust AI's judgment over that of one's own people. Leaders have been observed running team objections through chatbots, accepting the bot's verdict as the final decision. This not only stifles dissent but also reinforces AI's confident responses, as leaders receive less challenging input over time.
The lack of diverse perspectives allows AI's self-assured answers to appear increasingly plausible, creating a dangerous feedback loop. While delegation has been a common practice throughout history, the rapid adoption of AI and the ease of access present unprecedented challenges in maintaining clear governance and operationalizing its benefits.
Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
