There is a certain wildness in the tech industry these days that both mimics previous eras of large changes, like cloud computing with its runaway costs in the early days, and is like nothing we've ever seen before—record revenues accompanied by mass layoffs. One possible explanation gaining traction is that tech executives, especially CEOs, are collectively suffering from delusions of AI grandeur. Box founder Aaron Levie has publicly articulated this phenomenon, calling it 'AI psychosis.'

Origins of AI Psychosis

Levie, who has led Box through multiple technology cycles, argues that CEOs are uniquely prone to this condition because they are sufficiently distant from the last mile of work that still has to happen to generate most value with AI. When executives play with AI—whether by developing a prototype, generating a contract, or running a demo—they see the happy path results. They often do not consider the next 10 or 20 things that have to occur to turn that demo into a reliable, production-grade system. CEOs are not the people who review code, discover bugs, or identify calls to hallucinated libraries before software is deployed. They are not responsible for training AI models on a company's idiosyncratic contract terms, nor do they spend days combing through contracts to find sneaky terms. This lack of hands-on understanding creates a gap between perception and reality.

Levie, however, is no AI hater. He posts mostly AI-positive content on X to his 2.7 million followers, writing blogs such as 'Headless software is the future' and arguing that software built for AI agents is the way forward. He also backs AI startups as an active angel investor. Yet he recognizes that the executive suite's enthusiasm often outpaces actual capability.

Real-World Manifestations

The consequences of AI psychosis are already visible. In just the first five months of 2026, the tech industry has seen nearly as many layoffs as in all of 2025: 115,430 people have been fired from 152 tech companies so far in 2026, compared to 124,636 people let go by 275 companies in 2025, according to industry layoff tracker Layoffs.fyi. A bulk of these companies have pointed to AI as a reason for cutting jobs. Critics argue that many are AI washing—attributing cost reductions to AI when other business decisions are really driving the cuts.

One striking example is ClickUp, a project management and productivity software startup. CEO Zeb Evans proudly declared on X that he had laid off almost a quarter of his employees—22%—after rolling out about 3,000 AI agents to perform internal work. Evans swore this was not done to reduce costs. Instead, he envisions a workforce composed of people who run AI agents and spend their days quickly reviewing the agents' output. He calls this a '100x org.' But the data on AI and productivity does not support such assumptions.

Academic Perspectives

A meta-analysis of other research published in October in UC Berkeley's California Management Review found 'no robust relationship between AI adoption and aggregate productivity gain.' This suggests that despite the hype, the measurable impact of AI on overall output remains elusive. Research published in March by the National Bureau of Economic Research did conclude that AI adoption improved productivity but noted 'a productivity paradox, in which perceived productivity gains are larger than measured productivity gains.' In other words, companies feel more productive, but the numbers don't always back it up.

Further evidence comes from MIT, where researchers created thousands of agents to work on tasks. They concluded that agents are just not doing human-quality work yet in many cases. The MIT team predicted that at the current rate of LLM improvement, models will 'be able to complete most text-related tasks with success rates of, on average, 80%–95% by 2029 at a minimally sufficient quality level.' That means AI is on track to perform at base competence on most tasks in about three years, but agents will likely need another few years to outperform humans. For now, the gap between CEO expectations and actual AI performance is wide.

Meanwhile, research published in the Harvard Business Review highlighted an unintended consequence: when everyone in an organization uses AI to produce more stuff, the bottleneck simply shifts to executives. Their work now involves authorizing and overseeing all the output that AI enables. If employees are empowered to act independently, chaos can ensue—as OpenAI experienced last year. The result is organizational stress rather than streamlined operations.

The Risk of Organizational Chaos

The most certain outcome of ongoing CEO AI psychosis, if left unchecked, is organizational chaos. When top executives make sweeping decisions based on inflated beliefs about AI's capabilities, middle managers and frontline workers bear the brunt of unrealistic expectations. Mass layoffs based on overestimated AI productivity can erode institutional knowledge and morale. Moreover, the pressure to show results from AI investments may lead to cutting corners, such as deploying half-baked agents that require constant human oversight—something CEOs often fail to anticipate.

Levie's advice for CEOs is practical: use AI 'a ton' to truly understand what it can and cannot do. Only by experiencing both the upside and the real work—the bugs, the hallucinations, the edge cases—can executives come out the other side with an appreciation for both the potential and the limitations. But right now, too many CEOs seem to be acting on faith rather than evidence. The industry needs more leaders willing to publicly acknowledge the gap between AI hype and reality, and fewer who make sweeping cuts based on prototypes and demos.

Source: TechCrunch News