Record San Francisco house prices expose scale of AI wealth concentration

San Francisco has reclaimed its position as the most expensive city for homebuyers in the United States, with the median sale price hitting a record $1.76m in May. Year-on-year prices surged 19% in March, rising 14.5% in April and 14.1% in May, a trajectory entirely disconnected from the broader US market where the median home costs nearly $400,000.

Economists and real estate agents attribute this divergence almost entirely to the artificial intelligence sector. "They are just astronomical," says Daryl Fairweather, chief economist at Redfin. "People are flush with cash and ready to buy." The price spikes are concentrated in luxury Bay Area zip codes housing AI giants OpenAI and Anthropic, abruptly halting the city's pandemic-era property downturn.

The capital flooding the market is staggering. Last October, more than 600 current and former OpenAI employees sold shares worth $6.6bn, an average of $11m per participant. Anthropic workers recently executed a similar $6bn share sale. The wealth is so entrenched that some sellers are accepting AI company stock directly; a Duboce Triangle apartment recently sold for $3.2m after the seller indicated a willingness to take OpenAI or Anthropic equity.

For European investors and policymakers, this hyper-local inflation is a macroeconomic bellwether. It demonstrates that the AI boom is currently generating extreme, geographically concentrated wealth rather than broad economic gains. As OpenAI and Anthropic prepare for stock market flotations later this year or next, European institutional capital will likely pour into these listings, effectively financing a real estate boom on the US Pacific coast while Europe's own tech hubs see no comparable spillover.

However, the trajectory may not be infinite. Enrico Moretti, an economics professor at the University of California, Berkeley, says it is still "very early" in the AI boom, noting the city's employment and population remain below pre-pandemic levels. He also points out that the lion's share of upcoming IPO wealth will go to globally located investors rather than local employees. Broad tech layoffs at firms like Meta and the industry's eventual shift from innovation to established operations could also cool demand for highly paid, specialized staff.

In the interim, the housing market is aggressively pricing out the existing workforce. A non-tech government worker recently forced to relocate to a suburban town to afford a home captured the local tension: "We wouldn't have left if we could have afforded to stay. It kind of sucks and I do get a little salty seeing all this extra AI money squeeze everyone else out."