How Artificial Intelligence-Driven Models Could Help Prevent Forced Evictions in the US

Forced eviction has been a significant issue impacting low-income renters in the United States for many years. Facing the threat of eviction is a devastating experience for families who are already struggling to make ends meet, with many individuals unable to secure alternative housing in a timely manner. To help address this issue, researchers from the Penn State College of Information Sciences and Technology have developed two artificial intelligence-driven models that could help promote the rights of low-income renters in the US who are facing forced eviction.

According to data from the Eviction Lab at Princeton University, there were more than 900,000 evictions in the US in 2019, affecting over 2 million people. Renters from low-income backgrounds, particularly those from underrepresented communities, are disproportionately impacted by eviction, exacerbating existing societal issues related to income disparity, educational attainment, and mental health. However, despite the magnitude of this issue, the lack of accurate data and forecasting models has hindered progress towards finding a solution.

Weakly-Supervised Aid to Reduce Nationwide Eviction Rates

The first model developed by Penn State researchers, “Weakly-Supervised Aid to Relieve Nationwide Eviction Rate,” helps identify areas where there could be a high concentration of individuals facing eviction. The model utilizes machine learning techniques, including topic modeling, text analysis, and clustering algorithms, to scour publicly available data, such as social media, news articles, court records, and government reports, and extract relevant information to provide predictions. This model can inform decision-making and resource allocation to better address eviction rates across the country.

Multi-view model forecasting the number of tenants at risk of formal eviction

The second model, “Multi-view model forecasting the number of tenants at risk of formal eviction,” aims to provide an accurate forecast of tenants at risk of eviction at a certain point in the future. This model utilizes a combination of data sources, including municipal records, governmental reports, satellite imagery, and demographic data, to generate predictions. The model also takes into account the interactions between various factors such as local rent prices, employment rates, and policy changes that could impact eviction rates in the future.

Testing the models

To test the effectiveness of their models, Penn State researchers collaborated with the Child Poverty Action Lab and applied their models to a real-world dataset in Dallas County, Texas, where eviction records are more complete and readily available. The researchers compared the accuracy of their models against existing baseline models and found that both models outperformed some of the existing models by up to 36%. These findings are significant as they demonstrate the potential of these models in helping to prevent forced evictions.

Evaluation and Deployment

Both models are currently being evaluated by subject matter experts for a pilot deployment in the field. If successful, these models could be utilized by non-governmental organizations and policymakers to make more data-driven decisions about where to allocate resources to better address housing instability. The potential for AI-driven models to assist in addressing social issues goes beyond just housing instability, and these models could help lay the groundwork for more effective solutions in other areas.

The Penn State models represent a significant step toward using AI-powered technology to address housing instability and prevent forced evictions. While there is still much to be done to address this issue in the US, the use of these models offers policymakers more accurate and informed insights into the issue and its prevalence. If deployed effectively, these models have the potential to improve the lives of the millions of low-income renters across the country who are currently facing the threat of eviction. Dongwon Lee, a professor, and co-author of the study, sums up the importance of these models, saying, “Eviction disproportionately impacts individuals from underrepresented backgrounds and can exacerbate existing societal problems related to income disparity, educational attainment, and mental health. These models can help us better address these challenges and improve the lives of those vulnerable to eviction.”

Explore more

Can Recognition Solve Employee Engagement at Credit Unions?

A sincere and heartfelt expression of gratitude carries significant professional weight in a high-pressure lending environment, but it cannot offset the crushing weight of an unmanageable loan queue that grows every hour. While appreciation serves as a vital leadership tool, relying on it as a universal fix for morale risks ignoring the systemic pressures that staff face during the current

Building Strategic Recognition Systems for Modern Performance

When the pulse of a company is measured only by the quarterly bottom line, the vibrant heartbeat of individual effort often goes completely unnoticed. In the current landscape of 2026, the traditional model of patting an employee on the back once a year is not just outdated; it is a liability. This disconnect creates a vacuum where engagement withers, leading

Poland Emerges as a Premier Hub for E-commerce Fulfillment

The traditional logistics map of Europe is currently undergoing a radical transformation as major retailers shift their massive distribution centers eastward to capitalize on Poland’s unique strategic positioning. This evolution has turned the nation from a mere transit corridor into a sophisticated terminal for the most ambitious e-commerce brands on the continent. By 2026, the reliance on Central European hubs

Can Agentic AI Fix Modern Air Travel’s Support Crisis?

The modern traveler glides through terminal corridors with a digital boarding pass in hand, yet this seamless experience often shatters the moment a flight cancellation notification appears on their screen. This disconnect reveals a glaring invisible wall where the ease of modern booking apps ends and the archaic world of traditional customer support begins. While checking a bag or selecting

Managing the Rise of Customer AI Agents in Service

A consumer sits in a quiet living room while a personal AI agent navigates a complex labyrinth of “press one for billing” prompts, eventually securing a thirty percent discount on a homeowners insurance policy that the owner had intended to cancel for years. This scene, becoming common in 2026, represents a fundamental shift in the service economy where the traditional

Fill out the form below