Fourth of July weekend includes fireworks, hot dogs, parades, and parties.
For short-term rental platform Airbnb, these parties are no fun, but instead a disaster of noise complaints, injury risks, and property damage.
To combat this confusion and enforce the platform’s annual ban on open invitation parties, Airbnb five years ago began using machine learning technology to weed out what it called “high-risk” bookings. The technology will likely be working overtime during this weekend’s 250th anniversary celebrations.
The company says the technology uses “hundreds of signals” such as whether a user is booking on-site or for a short period of one or two days to indicate whether there are characteristics of a reservation that would turn the rental into a party house.
As a result, guests who are unable to book an entire home are instead redirected to book a private room listing or hotel on the Airbnb platform.
Last July 4, the company’s system deterred more than 20,000 people in the United States from booking entire residential properties on Airbnb, said Lucy Congo, a spokeswoman for the company.
He said about 2,500 people were directed in California, and about 200 were stopped, particularly in Los Angeles.
Airbnb said in a statement on its website: “These efforts reflect our continued commitment to helping disruptive parties reduce risk, and are seeing positive results.”
In 2025, fewer than 0.06% of Airbnb stays in the U.S. were reported as parties, the company said. Airbnb recorded 533 million booked seat nights in 2025, but did not surpass the number of real estate reservations in the U.S., according to a company filing.
Param Vir Singh, a professor of business technology and marketing at Carnegie Mellon University, believes the anti-party system is a useful tool for screening large bookings, given that the company “can’t have someone sitting down and evaluating every booking to see if it leads to a party.”
But Singh said machine learning tools that use large amounts of consumer data to make decisions can have blind spots and false positives. A case in point is Singh’s research on racial disparities in how hosts use the company’s free algorithm-based smart pricing tool to automatically set daily prices for accommodations.
Airbnb executives say they want to push these technologies even deeper and have a number of projects in the works.
During an earnings call in May, CEO Brian Chesky said that nearly 60% of the new code created by the company’s engineers is now written by AI.
“AI has a huge impact,” he said by phone.
Chesky is also in the early stages of funding a new AI lab he plans to set up separately from Airbnb, Bloomberg reported. According to Bloomberg, Chesky will remain Airbnb’s chief executive officer and will not play a central role in leading the new institute.
