AI Economic Disruption Scenarios

Scenario models where AI productivity gains trigger crises through success, not failure


Speculative but rigorous scenario modeling suggests that rapid AI capability improvements could trigger economic crises not through technical failure but through success — if productivity gains concentrate wealth while destroying the consumer spending that sustains the broader economy. The "2028 Global Intelligence Crisis" scenario illustrates how a positive feedback loop between AI investment, workforce displacement, and reduced consumer spending could produce a deflationary spiral.

The "Ghost GDP" scenario

Written as a fictional retrospective from June 2028, the CitriniResearch scenario models what happens when AI automation succeeds faster than the economy can adapt:

  1. AI capabilities improve rapidly (late 2025 onward) — agentic coding tools let small teams replicate mid-market SaaS products in weeks
  2. Companies cut white-collar workers — margins expand, earnings beat, stocks rally to S&P 8000 by October 2026
  3. "Ghost GDP" emerges — output shows up in national accounts but never circulates through the real economy; a single GPU cluster replaces 10,000 workers but spends nothing on discretionary goods
  4. Consumer economy withers — displaced workers spend less, velocity of money flatlines
  5. Negative feedback loop — margin pressure pushes more AI investment, which displaces more workers, which reduces spending further

Key mechanisms

Policy challenges

The scenario highlights how traditional policy responses fail:

Critical insight

The scenario's core thesis: AI-driven productivity gains are only economically beneficial if the resulting wealth circulates back through the consumer economy. When automation concentrates gains among capital owners while eliminating the income of consumers, nominal GDP growth masks a hollowing-out of actual economic activity.

— SOURCES
— GRAPH
— 4 RELATED