Wealthy AI workers send San Francisco house prices soaring
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- San Francisco's median house price hit a record $1.76m in May 2026, up 14.1% year-over-year, reclaiming the title of most expensive US homebuyer market from San Jose in March.
- Redfin chief economist Daryl Fairweather attributes the surge to AI money, noting steep luxury zip code price jumps since ChatGPT's late 2022 launch — a trend absent in cities with less AI wealth.
- A Duboce Triangle three-bedroom listed at ~$3m sold for $3.2m ($200k over asking), with the seller willing to accept OpenAI or Anthropic shares instead of cash.
- Last October, 600+ current and former OpenAI employees sold $6.6bn in shares (~$11m average per participant), while Anthropic workers recently sold ~$6bn in shares.
- Realtor Matthew Goulden reports bidding wars now routinely push sale prices millions over asking, with all-cash purchases surging particularly at the upper end of the market.
- A non-AI family whose parent holds a senior government job was priced out of SF to a suburban Bay Area town, while an OpenAI-connected family bought their SF home with an all-cash offer funded by company shares.
- Both OpenAI and Anthropic are due to hold full stock market flotations later this year or next, with Berkeley economist Enrico Moretti warning it is still "very early" in the AI boom.
Why it matters: With both OpenAI and Anthropic IPOs expected within roughly 18 months, SF's chronically supply-constrained housing market faces continued upward pressure from employees cashing in AI stock — forcing non-AI families, like one government employee's, to relocate 50 miles away while OpenAI-connected families buy in cash. The median SF home now costs roughly 4.4x the US median of ~$400,000.
