The San Francisco Bay Area has once again become the most expensive housing market in the United States in 2026, after home prices climbed throughout the spring.

The median sale price rose 19% in March from a year earlier, followed by increases of more than 14% in April and May. The average transaction price reached a record $1.76 million, far above the U.S. national average of roughly $400,000.

Daryl Fairweather, chief economist at the real-estate company Redfin, attributes the trend to buyers tied to the artificial-intelligence boom, who have “money to spend and are willing to buy.”

The number of homes for sale has fallen 40% from a year earlier, partly because owners are holding on to low-interest mortgages and because office-use patterns changed after the coronavirus period.

The squeeze has set off fierce competition for the properties that remain available.

At the same time, rents for two-bedroom apartments have risen 22%, reaching levels comparable to New York.

Local authorities and developers are even examining whether vacant downtown offices could be converted into housing to ease the shortage.

Real-estate agents describe a wave of cash offers and sales closing hundreds of thousands, and sometimes millions, of dollars above their original asking prices.

They link the dynamic to stock options held by employees of leading artificial-intelligence companies.

In the past month alone, 44 properties sold for at least $1 million above their asking prices.

Since the start of the year, there have been 144 such transactions.

One notable case involved a three-room luxury apartment in the Duboce Triangle neighborhood, initially listed at about $3 million and ultimately sold for $3.2 million after an OpenAI employee who toured it indicated a willingness to pay by transferring company shares.

The phenomenon is also being fueled by liquidity events among artificial-intelligence employees.

More than 600 OpenAI employees collectively sold $6.6 billion worth of shares in secondary transactions late last year, averaging about $11 million per person, while Anthropic employees sold roughly $6 billion in shares.

Analysts estimate that, following future public offerings, OpenAI employees could be able to buy about 20% of the homes sold in San Francisco, while Anthropic employees could purchase 9%.

As more new wealth flows into the high-end property market, ordinary residents say they are being pushed aside.

A San Francisco couple with a combined annual income of $365,000 said they struggled to find a one-bedroom apartment for less than $5,000 a month and felt unable to compete with technology workers tied to artificial intelligence.

Economists caution against assuming that the current trend will continue indefinitely, noting that the artificial-intelligence industry is still in its early stages.

Some companies within the sector are already developing their own housing policies to address the effects of the boom.

The startup Rilla, for example, spends about $1.7 million a year on housing assistance and allows employees to receive up to $18,000 annually to live near its Brooklyn office, as part of a broader benefits package worth about $37,000 per employee each year.