Artificial Intelligence Is Helping to Tackle Homelessness in LA by Identifying Housing Instability Before it Begins
The rapid growth of homelessness across Los Angeles County has inspired officials to create an innovative program that can help identify people at risk of losing their homes before it’s too late.
The model, known as the Homelessness Prevention Unit, was developed by California Policy Lab, the Los Angeles County Department of Health Services, the county’s Chief Executive Office, and the Department of Mental Health.
It works like this: Analysts use artificial intelligence to comb through 580 variables, ranging from enrollment in health services to usage of public benefits, contacts with the criminal legal system, and access to homeless services, to identify those most vulnerable to homelessness. Experts then conduct proactive outreach efforts to provide identified individuals with support or services to prevent them from becoming homeless.
HPU was launched at a time when homelessness has increased by roughly 43% in California since 2018. More than two-thirds of people identified as homeless reported that they were experiencing it for the first time. Meanwhile, rising rents and home prices continue to put additional pressure on low-income earning households.
“Solutions to homelessness have to include prevention,” said Janey Roundtree, director of the California Policy Lab at UCLA. “We have to reach people before they become homeless.”
Roundtree said the initial dataset experts compiled identified as many as 100,000 people in the state who could be at risk of experiencing homelessness. However, only 7% of that total was eligible for predictive services. Once they began filtering for people with high-risk factors, she said the number of eligible participants dropped to 10,000, of which 24% were considered high-risk.
Early Results Show AI’s Promise in Homelessness Prevention
Preliminary results from a test pool of 456 participants show that predictive methods could have a big impact on homelessness in California. So far, 92% of the participants completed the program, and another 86% retained housing since the program was launched in 2020. The model was also able to connect 42 people who were completely disconnected from California’s service ecosystem with the support they needed, Roundtree said.
A complete study of the program’s impact is expected to be released in 2027, Roundtree said.
According to Dana Vanderford, associate director of homelessness prevention at the Los Angeles County Department of Health Services, one key to the model’s success is its predictive services.
Those services include contact from homelessness prevention officials, either through phone calls, emails, or letters. Identified individuals are also matched with a case manager for four to six months and are offered flexible financial assistance options.
HPU staff also helps connect at-risk individuals with supportive services. That can include giving them a one-time cash payment to fix their car or pay rent. Vanderford described this approach as “measured but consistent.”
“Whatever our clients need, we try to make those dreams come true,” she added.
A Humane and Equitable Approach to Tackling Homelessness
HPU’s approach to using AI to address homelessness is radically different than other attempts. For instance, San Jose, California’s police department, created a first-of-its-kind AI tool that recognizes tens and cars where people experiencing homelessness are living. The police department drives a vehicle equipped with the technology through different parts of town to collect footage of streets and public places. This footage is then filtered through software like Ash Sensors, Sensen.AI, Xloop Digital, Blue Dome Technologies, and CityRover to determine what parts of town homeless people are living in.
Roundtree said the model is also equitable in how it predicts homelessness. That means it performs similarly for people of different ethnicities, heritages, and genders. This data can help people experiencing housing instability and service providers overcome discrimination and systemic racism that prevents individuals from accessing other services.
Going forward, the HPU team plans to work on retaining program staff by using private dollars to fund bonuses, skills training, and team-building activities. Reducing staff turnover is a key part of solving homelessness for several reasons. It helps service providers build rapport and trust with homeless people. It can also lead to more efficient and effective solutions by allowing staff members time to develop expertise in their jobs.
“By investing in the right proactive interventions, we’re not only preventing homelessness but also addressing challenges like mental health and economic stability—all of which ultimately save costs and lives,” Vanderfod said.
Advocacy and Innovation Must Work Together
Now is not the time to be silent about homelessness in California or anywhere else. Unhoused people deserve safe and sanitary housing just as much as those who can afford rent or mortgage.
Poverty and homelessness are both policy choices, not personal failures. That’s why we need you to contact your officials and tell them you support legislation that:
- Streamlines the development of affordable housing
- Reduces barriers for people experiencing homelessness to enter permanent housing
- Bolsters government response to homelessness
Together, we can solve homelessness.
