AI Psychosis: The New Leadership Blind Spot

The Rise of AI Psychosis in Executive Leadership

AI psychosis is a phenomenon where leaders develop an excessive, uncritical trust in AI-generated advice, often deferring to a chatbot's judgment over their own insights or the expertise of their human colleagues. While severe clinical cases remain rare, a milder version is increasingly prevalent in executive suites, where the perceived confidence and supportive tone of AI models lead leaders to loosen their grip on operational reality.

Recent data highlights the scale of this shift: 74% of executives report more confidence in AI's advice than in that of friends or colleagues, and 44% admit they would defer to AI's reasoning over their own insights. This trend is particularly dangerous because AI models are designed to be fluent and positive, often reinforcing a user's existing beliefs rather than challenging them with necessary friction.

Identifying the Symptoms of AI Over-Reliance

Excessive trust in AI manifests in three primary organizational symptoms:

1. The Production of "Workslop"

Leaders are increasingly failing to critically audit AI-generated content, leading to the proliferation of "workslop"—polished, fluent output that lacks substantive depth. Because AI drafts feel "finished" compared to rough human drafts, they are often sent to boards or employees after only a cursory skim.

This reliance has a measurable cost to leadership perception. Research from Stanford and BetterUp indicates that nearly half of employees view senders as less trustworthy, and over a third view them as less intelligent, when they receive AI-drafted emails.

2. Mandating Use Without Governance

Many organizations have mandated AI adoption as a baseline expectation before establishing the necessary guardrails. In some cases, managers are required to prove that AI cannot perform a task before they are permitted to hire a human. However, a 2026 Grant Thornton survey revealed that 78% of senior executives lack confidence that they could pass an independent audit of their AI governance within 90 days.

3. Deferring to AI over Human Expertise

The most critical symptom is the displacement of human judgment. Leaders may run team objections through a chatbot and allow the AI's verdict to settle internal arguments. This creates a self-reinforcing loop: as team members stop voicing objections to avoid being overruled by a bot, the leader hears only the AI's confident affirmations, further validating the leader's reliance on the tool.

Critical Perspectives on Cognitive Surrender

Industry practitioners and technical contributors suggest that "AI psychosis" may be better described as "cognitive surrender"—a choice rooted in laziness or a lack of critical thinking rather than a medical condition.

The Displacement of Institutional Knowledge

Technical contributors argue that C-suite reliance on LLMs ignores the nuance of actual implementation. As one observer noted:

"I think it's always been true that bad leadership tends to become out-of-touch with the reality of the business over time... with LLMs, that trend seems like it'll just get worse and worse."

This is particularly evident in software engineering, where a leader might trust a model's take on a codebase based on a small snippet, ignoring the years of institutional knowledge held by engineers who understand the "weird kludges" and technical debt of the actual system.

The Echo Chamber Effect

There is a psychological component to this trust; because users prompt the AI, they often implicitly guide the AI toward an answer that confirms their own bias. This transforms the interaction from seeking objective truth to seeking a sophisticated reflection of the leader's own existing opinions, effectively allowing leaders to defer to themselves via an AI intermediary.

The "White Collar Black Lung"

Some professionals report a behavioral shift in colleagues who use AI excessively, describing it as a state of "unreality" characterized by a manic, optimistic affect and a scattered, chaotic approach to work. This suggests that prolonged exposure to the "assured" voice of AI may degrade the organizational ability to maintain rigorous, structured professional standards.