A new scenario suggests that artificial intelligence could replace human labour on a scale and speed that the economy may struggle to manage, potentially sparking a “global intelligence crisis” by 2028. The prediction, framed as a fictional retrospective, examines how widespread AI adoption might reshape work, wages, and society.
The study, authored by James Van Geelen, chief executive of the American investment research firm Citrini, and AI entrepreneur Alap Shah, imagines a memo from June 2028 looking back at the crisis. According to the authors, large-scale layoffs began in 2026 as companies increasingly deployed AI agents to perform tasks without human supervision. Early 2026 tech-sector layoffs, including those announced by Amazon, Expedia, and Pinterest, are cited as foreshadowing the trend, though analysts caution it is difficult to directly attribute these cuts to AI alone.
The scenario outlines a self-reinforcing cycle: businesses adopt AI to stay competitive, AI systems become more capable, and companies justify further layoffs. “Each company’s individual response was rational. The collective result was catastrophic,” the authors note. Displaced white-collar workers are predicted to move into lower-paying jobs or blue-collar trades, while wage growth stagnates for those who retain employment. By mid-2027, the scenario projects the US economy tipping into recession, with unemployment exceeding 10 percent by 2028.
AI agents are envisioned as running continuously in people’s devices, performing complex tasks such as coding, long-term research, and financial optimisation. While AI creates new roles, including prompt engineers, safety researchers, and infrastructure technicians, these jobs pay substantially less than the positions they replace. The resulting oversupply of labour depresses wages further, forcing many households to rely on credit or retirement savings to meet living costs. By 2028, the scenario warns, the US could face conditions similar to the mortgage crisis of the previous decade.
The authors also foresee social unrest as part of the fallout. They predict a movement akin to the “Occupy” protests of the 2010s, with demonstrators in May 2028 blocking entrances to AI companies in Silicon Valley, including offices of Anthropic and OpenAI, to protest job losses and economic inequality.
Van Geelen and Shah caution that these outcomes are hypothetical and may not unfold exactly as described, but they argue the pace of AI adoption is already outstripping the ability of governments and institutions to respond effectively. The scenario serves as a warning that managing AI’s impact on employment, wages, and society will require urgent planning and policy measures if widespread economic disruption is to be avoided.
The report highlights the potential consequences of AI-related job displacement, noting that without intervention, automation could create a vicious cycle of rising unemployment, lower consumer spending, and social tension in the coming years.
