A pilot program in Los Angeles County uses predictive artificial intelligence to identify community members at risk of becoming homeless, and then intervenes to help them remain housed and stabilize resources.
According to a CNBC article, the California Policy Lab at UCLA has developed an AI model with existing data from local human service programs to identify individuals who might be at risk of experiencing homelessness. That information, linked together over many years, can predict individuals likely to experience homelessness. Armed with this information, the Los Angeles County Department of Health Services is able to reach out directly to individuals to offer support.
The data set included information from seven different county departments that was de-identified for privacy. Variables included emergency room visits, behavioral health care services, and large public benefits programs like food stamps, income support, homeless services, and the like. Through the program individuals and families were able to access rent and mortgage support in the range of $4,000 to $8,000, with no strings attached and without the need for an eviction notice. This has allowed intervention to happen before crisis strikes, which is often when recovery is even more challenging for these community members just on the cusp of resource stability.
