'AI Psychosis' Is the New Blind Spot for Business Leaders: When CEOs Trust AI More Than Real People
Fast Company points out that the 'AI psychosis' phenomenon is becoming a new blind spot for business leaders who rely too heavily on AI-generated answers, to the point of losing critical judgment in decision-making
The tech world is talking about a phenomenon called 'AI psychosis' from a new angle. A Fast Company article identifies it as the latest blind spot for organizational leaders, after many executives have come to rely on and trust AI-generated answers so much that they stop exercising their own judgment or verifying information themselves.
To clarify, AI psychosis is not a medical diagnosis, but a metaphor used to describe the behavior of business leaders. The term gained wider attention from a tweet by Mitchell Hashimoto, co-founder of HashiCorp, and was expanded upon by Aaron Levie, CEO of Box, who argues that CEOs are particularly prone to this condition because they are far removed from the 'last mile of work,' leading them to make decisions based on AI outputs that look better than reality.
Symptoms identified as 'AI psychosis' at the leadership level include placing greater confidence in AI's recommendations than in colleagues' opinions or one's own experience, failing to verify the accuracy of AI-generated work, and neglecting to ask questions—because AI tends to deliver answers that sound confident and reinforce the leader's existing beliefs, creating an echo chamber effect.
The article argues that what's happening is a leadership failure rather than a mental health issue, calling this state 'cognitive surrender'—surrendering one's thinking to AI.
However, the term AI psychosis is also used in another context in psychiatry, referring to cases where general users with psychological vulnerabilities may be triggered into delusional thinking or unhealthy attachment to AI. This is a separate issue from business use, despite sharing the same name.
The topic became a discussion on Hacker News with 150 points and 90 comments, reflecting that tech workers are also concerned about using AI without verification.
As Thai organizations begin adopting AI for decision-making, this issue serves as a warning that AI tools should not replace human analytical thinking—especially at the executive level, where flawed decisions affect the entire organization. It also raises questions about the need to verify AI-generated information before putting it into practice.