By Stuart Kerr, Technology Correspondent, LiveAIWire
The AI exodus from cities that many commentators predicted, low-skill workers in secondary metros priced out by automation while elite tech hubs thrive, is not what the data actually shows. Research from the Brookings Institution finds the opposite pattern: AI exposure is concentrated in exactly the cities assumed to be safest, the large, high-tech, well-educated metro areas including San Francisco, New York, Seattle, and Boston, because these places have the highest concentration of the white-collar, cognitive, nonroutine work that generative AI is best suited to automate.
That distinction matters enormously for how policymakers, workers, and city planners should think about where an AI exodus, if one happens, will actually land.
Which Cities Are Really Exposed to the AI Exodus
The occupations most vulnerable to AI substitution are not confined to factory floors or call centres in struggling regional towns. Coding, writing, financial analysis, marketing content, and presentation drafting, the tasks that large language models are best at, are disproportionately performed in the country’s largest, most economically successful metro areas. Brookings research on AI job exposure found that better-paid, better-educated professionals in bigger, high-tech metro areas face the most exposure to AI, a reversal of the pattern seen in earlier waves of robotics and software automation, which hit less-educated, lower-wage workers hardest.
A separate 2026 Brookings analysis, using occupation-level data on actual Claude usage, found that 62 of the 100 most AI-exposed counties in the United States voted Democratic in the 2024 election, driven by the concentration of office and information-based work, computer programming, marketing, financial analysis, in exactly the large urban counties that dominate those states. Rural and smaller communities, by contrast, showed meaningfully lower AI exposure across the board.
Why This Overturns the Standard Automation Story
The earlier wave of robotics and software automation followed a familiar geography: it hit manufacturing towns and lower-wage service workers hardest, while knowledge-economy hubs largely benefited. Generative AI is not following that script. Because it excels at the cognitive, nonroutine tasks that define white-collar office work, cities with a high concentration of high-earning professionals in tech, finance, law, and consulting are now the places facing the deepest occupational exposure, not the secondary cities that absorbed the last round of job losses.
This pattern echoes the broader cultural reckoning with AI-driven productivity that is reshaping expectations of work, place, and economic security across developed economies. The workers with the most to lose from this particular wave of automation are not the ones policy conversations have historically centred.
What Cities Are Doing, and Not Doing
As the AI exodus reshapes which cities actually face displacement, municipal governments in AI-exposed metros are responding with a mix of retraining programmes, economic development incentives, and housing policy reforms. The challenge is that the retraining pipeline is slow relative to the pace of displacement, and much of the existing policy infrastructure for automation response was built with manufacturing towns in mind, not knowledge workers in expensive coastal cities.
More promising are programmes focused on adjacent transitions, helping workers move into roles that AI augments rather than replaces. The parallel with the digital divide in access to AI tools is instructive: the benefits of technology tend to concentrate among those already advantaged, while the costs, in this instance, are landing on well-paid professionals in expensive cities who did not expect to be the ones most exposed.
The Political Dimension
The geographic concentration of AI exposure in Democratic-leaning, high-tech metro counties has become an emerging political fault line heading into the 2026 midterms. Brookings researchers are careful to note the correlation reflects occupational sorting, where knowledge workers cluster in cities that also vote Democratic, rather than a causal claim about ideology. But the practical effect is that anxiety about AI-driven job displacement is increasingly concentrated in exactly the places least accustomed to thinking of themselves as automation’s primary casualties.
For those navigating this landscape, the transformation of hiring and recruitment by AI systems adds another layer of friction to finding new roles, a compounding challenge for workers already under pressure from automation exposure in cities that were, until recently, assumed to be insulated from it.
What This Means Going Forward
The cities most likely to navigate the AI exodus successfully are those that recognise displacement as a structural policy challenge concentrated in their own knowledge-economy workforce, rather than assuming, as the earlier automation narrative suggested, that the risk belongs to somewhere else. That recognition requires city governments in San Francisco, New York, Seattle, and similar metros to build retraining and transition infrastructure aimed at their own white-collar workforce, not just the manufacturing and service roles that previous automation waves displaced.
The AI transition is moving faster than the industrial transitions of the twentieth century, and the geography of who it hits first has already surprised the people who assumed they were safest.
About the Author
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, cybersecurity, and the social impact of emerging technology. He publishes daily at LiveAIWire.com.