AI Boom Sends San Francisco Housing Market Soaring as Homes Sell $1M Above Asking Price

Wealthy AI workers send San Francisco house prices soaring

On a tree-lined street in the affluent Duboce Triangle residential neighbourhood of San Francisco, the top half of a white, Edwardian-era, detached house was drawing visitors from prospective buyers. The opulently renovated three-bedroom apartment was on the market for almost $3m (£2.3m). And it had been attracting increased attention due to an unusual payment possibility - the seller would consider shares in artificial intelligence companies OpenAI or Anthropic instead of cash. "The value [of the property] is questionable, but I would like to buy," says a young OpenAI employee who has just viewed the flat with his partner. The worker, who moved to the Californian city two years ago for a technical job with the San Francisco-based company, is currently renting. He plans, he says, to ask his bosses about the stock transfer possibility. Welcome to San Francisco 2026, also home to fellow AI giant Anthropic. The city is ground zero for the AI revolution, and its property prices have risen dramatically this year. "They are just astronomical," says Daryl Fairweather, chief economist at Redfin, a real estate company that tracks US home prices. "People are flush with cash and ready to buy." In March, San Francisco regained its title as the most expensive city for homebuyers in the US, overtaking rival San Jose 50 miles to the south in the heart of traditional Silicon Valley. That month, the median house price in San Francisco rose 19% on the year before, and that trend has continued, up 14.5% and 14.1% in April and May respectively, according to data provided by Redfin. The median sale price in the city as of May 2026 is a record high of $1.76m, compared with nearly $400,000 for the US as a whole, where prices rose by just 1.4% in March, and 2% in both April and May. The prevailing view of pretty much everyone is that AI money is the driver of the red-hot San Francisco property market. "We have come to that conclusion based on what we're seeing in the data, and what we've heard from our agents," says Fairweather. She highlights the steep jump in prices in the wider San Francisco Bay Area's luxury zip codes - which includes Duboce Triangle - since OpenAI launched ChatGPT in late 2022, a trend absent in cities with less AI wealth. It has halted the downturn that San Francisco saw during the Covid pandemic, when the population fell and house prices softened. Today, the high salaries and signing bonuses being paid to top AI staff in the city can be extraordinary, even by Silicon Valley standards. Yet even more generous are the stock options that the employees have been allowed to partially cash in via limited share sales. Last October, more than 600 current and former OpenAI employees sold combined shares worth $6.6bn, an average of $11m per participant, it was recently reported. At Anthropic, whose main product is Claude, workers were also recently said to have been allowed to sell shares totalling some $6bn. And with both companies due to have full stock market flotations later this year or next, minting more multi-millionaire employees, many see no end in sight to San Francisco's real estate rises. "Today's bidding wars are going to be seen as bargains, and they already are," says Rachel Swann, the listing agent for the Duboce Triangle property. Enrico Moretti is a professor of economics at the University of California, Berkeley, who lives in the city. He says it is still "very early" in the AI boom, and points out that while the city's population and employment levels are rising, they remain below what they were before the pandemic. There are also opposing forces that may keep a lid on things. Big tech firms such as Meta have recently seen large layoffs. And as the AI industry moves from its fast-growing innovation phase to one of established companies, it is likely to require less specialized workers who are less able to command the same pay. Moretti also points out that the lion's share of the wealth from OpenAI and Anthropic's coming stock market flotations will go to investors rather than employees, and they are globally located. But in the meantime, San Francisco estate agent Matthew Goulden says the current situation is "crazy". Goulden, who has been doing the job for more than 20 years, says he first started noticing an uptick in prospective buyers - many from the world of AI - late last year. The upward trend, he says, is not just confined to luxury properties but extends across the market, from single-family homes to one-bedroom flats, and while it is most pronounced in desirable neighbourhoods, it is being felt almost everywhere. He says that bidding wars are now common, sometimes pushing sale prices millions above the asking level. At the same time, he adds that homes are selling faster than ever, and the number of all-cash purchases seems to be surging, particularly at the upper end of the market. Danielle Lazier, another experienced San Francisco realtor, describes similar, but adds some perspective. There has long been a tendency in San Francisco for homes to be listed below market value to get an auction effect going, she says. And supply is chronically limited - San Francisco is small, there is a high proportion of renters and it has struggled to build new housing (even if the city's new pro-growth, recovery-focused mayor is seeking to change that). "All of a sudden AI money can have an outsized effect," she says. Meanwhile, as the new AI boom takes hold, the tale of who gets to stay in San Francisco and who doesn't is told by its residents. Two San Francisco families with school-aged children, who both asked for anonymity to protect their privacy, recently succeeded in buying move-in-ready single-family homes to meet their desperate needs for more space - but only one was able to do so in the city. That family was able to purchase in the desirable family-friendly neighbourhood where they had been long-term renters after one parent, who works at OpenAI, sold some company shares last October, giving the family the financial boost needed to buy in an all-cash offer. The couple say they feel "conflicted and self-conscious" that it is AI money that has made it possible. "We're not ostentatious people," they add. "We've just done what we can with the opportunity." In contrast, the other family, which doesn't derive its income from AI or the tech world, had to instead move to a more suburban Bay Area town to the north. Their new home, bought in part with a mortgage, includes a pool and extra land. It is a different kind of life, notes the mother, and they have mostly adapted now - though it involves a long commute for her husband, who has a senior government job in San Francisco, and they still have "what if" moments. "We wouldn't have left if we could have afforded to stay," she reflects. "It kind of sucks and I do get a little salty seeing all this extra AI money squeeze everyone else out." The Duboce Triangle flat, for the record, and according to its listing agent, sold for $3.2m - $200,000 over the asking price. Whether the deal included AI stock is confidential.

In San Francisco, Some Home Sellers Now Ask for OpenAI or Anthropic Stock

When Nima Gabbay decided to sell his three-bedroom, two-bath San Francisco home for $2.995 million last month, his listing described the residence's soaring 10-foot ceilings, kitchen wrapped in Calacatta marble, remote-control skylights and oversize two-car garage. The 51-year-old real estate investor and developer also added an unusual clause: He would accept shares of OpenAI or Anthropic as payment for the home. Two OpenAI employees soon came forward offering some of their shares for the property, Mr. Gabbay said. One bid more than $1 million above the asking price, but appeared to inflate the value of his OpenAI stock. The other backed off when OpenAI filed to go public last month, deciding to hang on to the stock. Mr. Gabbay ultimately went with a third buyer who works in tech, and the sale is set to close this week. He was not at liberty to disclose the sale terms or the buyer's identity because he had signed a nondisclosure agreement, he said. "There's a bit of a gold rush situation right now in San Francisco," Mr. Gabbay said. Selling the home was "an avenue for me to potentially pick up some of this stock and be a part of the excitement of the companies going public." Even before OpenAI and Anthropic hold initial public offerings, the artificial intelligence companies -- which are based in San Francisco and leading the A.I. boom -- are distorting the city's housing market. Sellers are asking for pre-I.P.O. stock as payment for homes, property prices are surging as buyers bet that whatever they overpay today will look cheap tomorrow, and landlords are pushing out tenants to sell into the hotter market. The maneuvering is aimed at getting ahead of the wave of wealth when OpenAI and Anthropic, each valued at nearly $1 trillion, go public. Their I.P.O.s, plus the recent public offering of Elon Musk's SpaceX, could create more than 16,000 millionaires and more than 20 billionaires, according to Sacra, a private market research company. Already, San Francisco's sales of homes above $10 million have doubled over the past six months compared with a year earlier, said Joel Goodrich, an agent with Coldwell Banker Global Luxury. Forty-four homes closed at prices at least $1 million more than their asking price last month, said Mike Simonsen, the chief economist for Compass Real Estate. And there have been 144 such sales so far this year, up from eight in the first half of 2025. Fewer than 600 homes -- including single-family houses and condos -- are on the market today, about 40 percent below San Francisco's average of the past decade, according to Compass. The market is so frenzied that a six-bedroom, seven-bathroom 5,725-square-foot home in the Cow Hollow neighborhood with views of the Golden Gate Bridge and Alcatraz sold for $15 million in May, nearly double the list price of $7.9 million, according to John Caruso of Sotheby's International Realty. Even in a city that lived through the late-1990s dot-com boom and major public offerings by companies like Google (in 2004), Facebook (2012) and Uber (2019), property agents and wealth managers said they had never seen anything quite like this. "There's a hysteria that's out there right now," said Pete Rodway, a Compass agent who works mostly in the luxury market. One of his clients, an OpenAI employee, was scrambling to buy a $5 million home now to beat "a thousand other people that are going to have a budget of $30 million," he said. Garret Spiecker, who works at Citizens Private Bank and describes himself as a "financial therapist" for sudden wealth, said he had advised dozens of OpenAI and Anthropic employees on how to navigate the housing market. He has suggested they buy properties through trusts to protect their privacy, especially on homes above $5 million. "In this cycle, which differs from some of the others, a lot of these individuals are very young and quite wealthy very fast," he said. Anthropic and OpenAI, which have not set dates for their I.P.O.s, declined to comment. (The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The companies have denied the claims.) Techies who do not work at OpenAI or Anthropic have accelerated their plans to buy homes. Sam Rosenstein, 31, a software engineer at the software company Databricks, and his partner, Michelle Huang, 31, who works in tech sales, jumped into the market this spring partly because they wanted to close on a deal before the flood of A.I. wealth, Ms. Huang said. "There was just a general acceptance that that time will eventually come," she said. Their urgency spiked when Mr. Rosenstein's landlord decided to sell his rental property to cash in on the rising market. But the competition for a home was so fierce that one house the couple bid $600,000 over asking for in April ended up selling for roughly $900,000 more than its asking price. In May, Mr. Rosenstein and Ms. Huang landed a four-bedroom, two-bathroom in the Hayes Valley neighborhood for $2.185 million, bidding $385,000 above asking. "When we put in the offer for the house, the seller came back and said we could pay the offer that we put in, or we could pay less money but provide 60 hours of A.I. consulting" on a personal project, Mr. Rosenstein said. "That's totally the weirdest thing that has happened." They declined the $10,000 discount. Like Mr. Gabbay, other home sellers are unabashedly angling for shares of OpenAI and Anthropic. In April, Storm Duncan, 56, the founder of the tech-focused investment bank Ignatious, quietly marketed his 4,372-square-foot four-bedroom, five-bath compound in nearby Mill Valley, Calif. -- featuring an infinity pool and views of the San Francisco skyline -- on a LinkedIn page he created just for his house, which he valued at approximately $8 million. He direct-messaged Anthropic employees and investors, hoping to trade the home for stock. The listing went viral after someone from Khosla Ventures, a venture capital firm that has invested in Anthropic and OpenAI, leaked the LinkedIn post, Mr. Duncan said. The California Post published an article about the property soon after. Mr. Duncan took the listing down, though he said he would still do the deal if the right opportunity arose. Anthropic is "narrowly focused on building a great product," he said. In May, Vijay Chattha, 49, a tech entrepreneur, listed his three-bedroom vacation home in Sonoma County wine country, an hour's drive from San Francisco, with a $500,000 discount off the $2.5 million price if the buyer paid in Anthropic stock. "I think Anthropic is going to grow faster than the real estate market, so why not just do a trade?" Mr. Chattha said. He added that he already had OpenAI stock and wanted to use the deal to build a stake in Anthropic. The vacation home, now listed at $2.35 million, has not sold. But Mr. Chattha said he was undeterred. He next plans to list a condo in San Francisco -- also for A.I. stock.

How AI wealth is reshaping San Francisco's housing market

Why it matters: It's a striking illustration of how much purchasing power could enter a housing market already defined by scarce inventory and fierce competition. By the numbers: Former and current OpenAI employees could theoretically buy about a fifth of all residential real estate in the San Francisco metro, based on Redfin's estimate of roughly $135 billion in post-tax equity. * Anthropic employees could account for another 9% of the market with an estimated $63 billion. * Combined, that's enough buying power to cover about 29% of the metro's $692 billion housing market, based on Redfin's estimate of the total value of all homes in 2024. The big picture: Both companies could go public with reported valuations around $1 trillion. * Local home prices are already climbing at their fastest pace in years as highly paid AI workers compete for a limited number of listings. * Some properties are selling for more than $1 million over asking as buyers race to get into the market before the anticipated AI windfall arrives, the Wall Street Journal reports. Between the lines: Some OpenAI and Anthropic employees have already unlocked some of their wealth through secondary share sales, company tender offers and loans backed by private stock, giving them cash to buy homes before an IPO, per the New York Times. * That's a departure from previous tech booms, when employees typically had to wait months after an IPO for lockup periods to expire before accessing their equity, San Francisco real estate agent Karen Mendelsohn Gould tells Axios. * "This is one of the first times this has ever happened,"she says. "It was a big surprise that they were going to be able to get access to their money before the companies went public." The intrigue: The frenzy has become so intense that some home sellers are marketing properties with an unusual pitch: they'll consider accepting pre-IPO stock as part of the purchase price. Yes, but: Don't expect that to become the norm, real estate broker Danielle Lazier tells Axios. * "It's hard to imagine more than a handful of outlier sellers wanting stock shares over cold hard cash," Lazier says, comparing it to the short-lived push to buy homes with Bitcoin. Plus, those deals remain rare and face significant hurdles because restricted stock can't be transferred before it vests or without company approval, Mendelsohn Gould says. The bottom line: In real estate, the expectation of AI wealth may be proving almost as powerful as the wealth itself.

'Absolutely bananas': San Francisco homes sell for $1m above asking price amid AI boom

Report finds widespread overbidding as rapid AI growth generates increased wealth in city where housing is scarce San Francisco's AI boom has buyers spending unprecedented amounts of money on homes - much more than sellers are asking for. A new analysis from real-estate brokerage Compass, found that in the first half of 2026, more than 140 homes in the city sold for at least $1m above their asking price, 44 of them in June alone. The figure is a huge jump from January through July of last year, in which only eight homes sold for more than $1m, according to Compass data. Just six homes sold above that amount in the first six months of 2024. The skyrocketing demand for homes in this price range is "absolutely BANANAS", according to Mike Simonsen, Compass's chief economist. Simonsen has previously explained that the widespread overbidding was "of course related to the AI boom. It's migration and hiring, as well as preparing for mega IPOs." OpenAI and Anthropic, both headquartered in San Francisco, have filed to go public on the US stock market at valuations approaching $1tn. Their debuts promise to mint a new class of multimillionaires in San Francisco, which is already home to the highest concentration of billionaires per capita in the world. San Francisco's single-family home prices have risen about 17% year over year, while inventory has plummeted roughly 45%, according to Compass's market intelligence report published last month. The single-family median home price has increased from $1.7m to $2.2m, and houses are getting snapped up at an average of 18 days on the market, the fastest pace in five years. The sky-high sales in San Francisco's housing market marks a shift from a few years ago when departures as well as concerns about crime and homelessness contributed to a slump. Compass's report noted that in San Francisco, "AI and tech-driven demand has created aggressive bidding wars on the scarce inventory" and that "skyrocketing rents are back in the norm". Simonsen explained that the resurgent demand is so far concentrated "to a very small section in the city, and luxury markets in Peninsula and Marin". The report describes a "housing market increasingly segmented by income tier and proximity to AI-driven employment centers". San Francisco stands out as having the highest median home price in the country, according to a May 2026 analysis from real estate firm Redfin. Simonsen also said it was notable that other tech hubs across the country have not seen a similar trend of overbidding. Redfin found that San Francisco had the highest nationwide increase in the median sales price of homes from a year earlier, followed by Detroit and Providence, Rhode Island. The city reported a more than 10% increase in April, compared to the prior year. "What's different this time is that the benefits or the prosperity of AI seems much more concentrated," said Daryl Fairweather, the chief economist at Redfin, according to the New York Times. "It's not that everybody is going out and buying homes."