Los Angeles uses AI to fight homelessness crisis

Homeless encampments line the bike path on Venice Beach in Los Angeles, California, U.S., April 14, 2021. REUTERS/Lucy Nicholson

What’s the context?

In LA County, AI pinpoints people at risk of homelessness in a ground-breaking pilot that shows how tech can be used for good

1. Record numbers of people without homes in the United States
2. LA County's AI tool highlights vulnerable individuals
3. Participants in a pilot program receive cash and guidance.

WASHINGTON - In June of last year, Breanna Sanchez found herself in a difficult situation. Her Los Angeles landlord was threatening to evict her family because she was late on the rent.

However, an unseen algorithm helped toss her a lifeline.

An artificial intelligence (AI) technique utilized as part of Los Angeles County's novel experiment to identify individuals at risk of losing their houses within a year had identified Sanchez's precarious predicament.

The tool generates a list of 100,000 persons judged to be at danger of ending up on the streets by analyzing more than 400 indicators, including information from the county jail and hospitals, safety net programs, and the systems for homelessness and foster care.

Then, in an effort to prevent more people from becoming statistics in the rapidly expanding national homelessness epidemic, city officials set out to provide advice and even financial support.

Sanchez received a letter advising her to seek housing assistance from a county department. The 33-year-old soon began speaking with an outreach worker, registering for health and other programs, and obtaining financial aid to help with rent, car repairs, and other expenses.

"I think I would have been kicked out," Sanchez said to Context.

"Me and my husband felt stuck and didn't know how to move forward."

Smaller trials are also being conducted elsewhere as authorities deal with a homelessness problem that has forced some cities, including Los Angeles, to declare states of emergency. The Los Angeles pilot program started in 2021.

According to Dana Vanderford, associate director of homelessness prevention for the county's health services department, roughly 87% of the hundreds of individuals the new L.A. Homeless Prevention Unit has worked with have kept their homes after leaving the program.

"They frequently say, 'I don't know how you found me,' when we phone them during a crisis. I don't know what to do because I'm losing my apartment next week," Vanderford remarked.

"Our ability to appear out of nowhere and intervene and resolve a crisis - we're really proud of that," she continued.

AI is also being used in Calgary, Canada, to help anticipate the danger of homelessness, and according to Geoffrey Messier, who is assisting in the project's development, the L.A. pilot might be a game-changer.

"I hope it is a bit of a watershed moment," said Messier, an electrical and software engineering professor at the University of Calgary. "This is the first time machine-learning was an integral part in identifying people who needed help."

'Right way to apply technology'

U.S. Housing and Urban Development reported that in 2023, there were over 653,000 homeless persons in the country, a 12% rise from the year before.

Recent research, according to Janey Rountree, executive director of the California Policy Lab at the University of California, Los Angeles, suggests that additional services in addition to a one-time cash payout could help halt that trend.

In a research released last year by the University of Notre Dame, it was discovered that within six months of enrolling in a financial assistance program, at-risk individuals in Santa Clara County, California, had an 81% lower chance of becoming homeless.

"You can put the household back on a trajectory toward housing stability if you get them at the right moment and give them $2,000 to $6,000 of assistance to help pay debt or rent," stated Rountree, who is involved in the L.A. pilot's evaluation and helped lead the research that led to its creation.

The new AI pilot will help local officials locate persons in need, but first they must be able to identify them.

Case managers like Jocelyn Bataz of the L.A. Homeless Prevention Unit get to work when someone is identified as being at risk. 

She explained that some people are first dubious, saying, "I call them out of the blue and introduce myself." She tries to find out about their housing status, utility bills, and other worries after assuring them that the contact is genuine.

Homeless Prevention Unit programme participant Breanna Alvarez, left, and case manager Jocelyn Bataz in March 2024. Heidi de Marco/Los Angeles County Department of Health Services/Handout via Thomson Reuters Foundation.

Those who qualify can receive financial aid and other services in a matter of days, such as helping them pay off debt, locate childcare or employment, and obtain insurance. In addition, Bataz claimed to offer substantial emotional support.

"To ensure that we're reaching people who are at risk but aren't otherwise connected to homeless systems, this is the right way to apply technology," stated Alex Visotzky, a senior policy fellow with the National Alliance to End Homelessness based in California.

Eye of AI

According to Harvard University urban policy expert Stephen Goldsmith, AI technologies can be more efficient than manual methods in determining who is most chronically in need of housing support.

"Because preventing (homelessness) is complex, AI tools can help cities be more effective by focusing necessary services on specific individuals and their needs," he stated.

AI is also being utilized in London, Ontario, a Canadian city where the COVID-19 outbreak caused a twofold increase in the homeless population.

Director of London's Housing Stability Services Craig Cooper claimed that an artificial intelligence tool is being used to analyze data from those currently in the shelter system in order to determine who is most vulnerable. The model looks to be rather accurate thus far.

In Calgary, similar initiatives are being carried out to address the issue of newcomers to the shelter system usually having to wait a long period to be eligible for housing aid. However, AI can analyze early system data to evaluate longer-term threats.

According to Messier of the University of Calgary, "if so, that increases their priority for support."consumed.

Only a few months after finishing the program, Sanchez is thankful to have her life back on track in Los Angeles. She is expecting a child, her health issues are being handled, and she will graduate from high school in June.

Regarding the pilot, "it really helped a lot," she stated. "I feel safer here now."

(Editing by Clar Ni Chonghaile, reporting by Carey L. Biron @clbtea.)