A massive influx of AI workers is jacking (opens in new tab) up rents in San Francisco. They’re crowding open houses, paying for a year upfront, and dousing (opens in new tab) the market with newfound cash.
At least, that’s what the headlines suggest.
But how real is that housing boom? And what does the data mean for renters? A review of several rental sites shows that prices aren’t rocketing the way some reports have suggested.
Instead, the rise may feel steep as San Francisco creeps back to pre-pandemic rent prices — and the AI narrative may be adding fuel to that little fire.
In New York City, an apartment that rented for $2,500 before the pandemic is now likely to cost more than $3,200, according to four datasets analyzed by The Standard. In San Francisco, that apartment would still cost around the same as it did in 2020.
Zillow’s data shows that SF rents have risen 4% since the eve of the pandemic. In that same period, Angelenos have seen a nearly 23% rise, and New Yorkers have weathered a 29% jump.
But any increase is relative, and the return to pre-pandemic prices may shock San Franciscans who had grown accustomed to the unusual discounts of the last few years.
San Francisco’s sluggish rebound may be partially due to workers who left during the pandemic. Now that has reversed: Economists say return-to-office mandates, AI companies included, might be fueling overdue increases in residential rent prices. Indeed, some data suggests (opens in new tab) that more people are looking to rent in San Francisco than a year ago.
But whether you’re a boom believer may depend on which listing site you’re looking at.
If you got your numbers only from Apartment List, you’d probably think the boom is real.
The site estimates that prices across all unit types rose 11% over the last year. Rob Warnock, head of research at Apartment List, said his platform tends to see more listings for large, newer, professionally managed apartment complexes — the type where new-to-town young AI workers may be flocking.
Contrast that with Zillow, which has five times as many listings as Apartment List in San Francisco and generally sees more diversity in them. It has the most conservative year-over-year estimate, at a little over 3%. That’s higher than the national average and a percentage point higher than the regional inflation rate, but equal to typical pre-pandemic rent increases.
CoStar, Redfin, and Zumper split the difference, with rises estimated at 6%, 7%, and 8%, respectively.
“It’s a good idea to look at multiple sources,” said Daryl Fairweather, Redfin’s chief economist. “If one is telling a wildly different story than the others, then I think there’s reason to be skeptical.”
Every rental site has its own method of estimating prices. Some use sophisticated calculations meant to account for various biases, or pull additional data from outside sources.
Ultimately, though, each site analyzes a different set of numbers in a different way and can come to a different conclusion about what the market is doing. Case in point: Zumper publishes rent prices that are consistently much higher than those on other platforms.
Just this week, its economists reported (opens in new tab) that two-bedrooms in San Francisco have reached their highest rate, $5,000, in the company’s decade-long tracking history. Rival companies have not reported the same rise in rents.
When asked about the discrepancy, Crystal Chen, a spokesperson for Zumper, pointed to differences in its inventory and calculations. “It really comes down to methodology,” she said.
Zumper says (opens in new tab) it reports the median of current asking rents, versus the more complex calculations made by Zillow (opens in new tab) and Apartment List (opens in new tab).
Outlier data points are more likely to make headlines than modest estimates — and they can become a self-fulfilling prophecy when landlords are looking for reasons to set higher rent prices. Combined with real market forces like higher demand (opens in new tab) amid persistent lack of supply, the feeling that rents are already swiftly on the rise can quickly turn into real hikes.
Warnock speculated that large multifamily operators are taking advantage of the buzz around the AI industry to raise rents, though he cautioned that that’s difficult to quantify.
“Big professional landlords who set rents have been waiting for this opportunity to push prices up again,” Warnock said. “I think there is an appetite for raising rates to try to recuperate everything that was lost.”
Research supports this theory. Studies by economists at Copenhagen Business School (opens in new tab) and the University of Michigan (opens in new tab) found that news sentiment can significantly influence prices in the U.S. housing market. Further (opens in new tab) research suggests that the effect is most pronounced when homebuyers and renters have less data available to them.
“The information (or misinformation) disseminated through news media” can have a “persistent impact on price discovery and market participant behavior,” economist Oguzhan Cepni wrote in the Copenhagen Business School study.
That is to say: If everybody believes that rental prices are going up, they probably will. We might all be implicated in this particular AI hallucination.