On February 22, 2026, macro research firm Citrini Research published a provocative thought experiment titled The 2028 Global Intelligence Crisis. Written in collaboration with Alap Shah, the report is framed as a future-dated macro memo from June 2028. Its premise is not that artificial intelligence fails. It is that AI works better than expected and that very success destabilizes the global economy.
The document is explicitly presented as a scenario analysis, not a prediction. The authors describe it as a way to model a left-tail risk that markets may be underexploring. At a time when investors largely view AI as a margin-expansion and productivity story, the report asks a deeper macroeconomic question: what happens if intelligence becomes abundant faster than income can adjust?
The report imagines a world between 2025 and 2027 in which AI-driven productivity surges. Companies aggressively adopt autonomous systems, replacing large segments of white-collar labor. Corporate margins expand dramatically. Earnings beat expectations. Equity markets rally. Capital expenditures shift heavily toward compute infrastructure.
On paper, the economy looks strong. But beneath the headline growth, the authors introduce a critical concept they call Ghost GDP. This refers to economic output generated by machines that shows up in national accounts but does not circulate through the human economy. AI agents do not consume, borrow, form households, or spend discretionary income. As more output is produced by non-consuming entities, aggregate demand weakens even as productivity rises.
The scenario’s fictional June 2028 memo describes unemployment rising above 10 percent and the S&P 500 down nearly 40 percent from its peak two years earlier. Markets initially celebrated efficiency gains, but over time the erosion of household income began to undermine consumption, credit quality and asset prices.
At the center of the analysis is what the authors call the Human Intelligence Displacement Spiral. The mechanism unfolds in stages. AI systems improve rapidly. Firms reduce headcount in knowledge work. High-income workers lose earnings power. Consumer spending declines. Revenue pressures build. Companies respond by accelerating AI adoption to defend margins. AI improves further, and the cycle intensifies.
Unlike past automation waves that primarily affected manual labor, this scenario targets high-income knowledge workers — the same demographic that underpins mortgage markets, discretionary spending and urban real estate demand. Because these workers are deeply embedded in credit markets and asset valuations, their displacement creates ripple effects across housing finance, private equity-backed software firms and venture portfolios.
The report also argues that AI erodes business models built on cognitive friction. Entire sectors that monetize coordination, synthesis, routing, negotiation or analysis face compression as autonomous agents reduce marginal costs toward zero. Software products once viewed as defensible become reproducible. Service layers once considered essential become optional.
A key concern in the report is institutional speed. Technological change accelerates exponentially, while policy adapts more slowly. In the imagined 2028 environment, redistribution mechanisms, retraining programs and fiscal responses lag behind labor displacement. The resulting gap amplifies deflationary pressures and erodes confidence.
Importantly, the authors do not claim collapse is inevitable. The purpose of the February 22 publication was to stress-test assumptions embedded in today’s AI optimism. Historically, productivity revolutions have ultimately expanded employment and living standards. The report challenges whether the pace and concentration of modern AI gains could produce a more disruptive transition.
Its central question is straightforward but profound: if machines produce an increasing share of economic output while humans earn a decreasing share of income, who sustains aggregate demand?
For investors, The 2028 Global Intelligence Crisis reframes AI from a pure earnings story to a distributional and macroeconomic story. It suggests that productivity gains alone do not guarantee stable growth if purchasing power becomes increasingly detached from production.
The AI boom may yet deliver extraordinary prosperity. But as Citrini Research’s February 22 thought experiment makes clear, prosperity depends not only on how much intelligence exists in the system, but on how its gains are distributed and whether economic institutions can evolve quickly enough to absorb the shock.