The housing industry is growing less productive at precisely the time we need it to do more, making it harder and more expensive to meet the nation’s housing needs. Artificial intelligence, with humans making the decisions, can help.
The rapid evolution of AI and the way it is being used raise serious questions about how it’s shifting the economy, as well as the environmental and energy impact of the infrastructure needed to run it. But as we contend with a national housing crisis, it’s important to recognize that we can put its ability to create efficiencies to work.
Used responsibly, AI could help improve the housing industry’s productivity, which has been falling off since the 1970s, speed up housing production and preservation, and better target services to keep Americans from becoming homeless.
Of course, AI will not build a single home on its own, and it should not be trusted to decide where money goes or who gets housed. People still must make those judgment calls. But urban planners and others in the housing industry are experimenting with AI to collect, organize, analyze and process the enormous amount of data and information that helps inform those decisions.
How data tools can clear pre-development hurdles
The Urban Institute is examining how AI can help planners more quickly analyze complex zoning information – work that could ultimately help communities identify where more housing can be built near transit and other amenities while minimizing displacement of existing residents.
As these tools develop, they could reduce the time planners and researchers spend manually gathering and interpreting what can be built and where, while leaving the final planning and land-use decisions to people.
In Seattle and Honolulu, AI is already chipping away at the permitting bottleneck that hampers development. Seattle piloted an AI screening tool to catch missing or incomplete information in construction permit applications before they reach reviewers, potentially cutting intake review time by roughly half and reducing correction cycles by about 35%. In Honolulu, residential projects using AI-powered CivCheck reached a permit decision in an average of 32.5 days in early 2026, down from 73 days, while average corrections fell from 23.5 to 7.7.
Those gains matter. Every unnecessary month spent navigating permitting, correcting paperwork or reconciling information can add costs to a housing project, and in an already difficult financing environment, delay can be the difference between a project moving forward or not.
Lowering building costs and preserving vulnerable homes
AI can also help us reduce building costs that are surging at a rate not seen since the COVID-19 pandemic. We’re already seeing companies deploy machine learning, robotics and automation to operate microfactories close to building sites that can deliver homes with increasing efficiency. One company says its process can cut construction costs by a third and deliver homes three times faster than traditional construction.
The same principle applies to preserving the affordable housing we already have. The earlier communities can identify properties at risk – because of financial distress, deferred maintenance or expiring affordability restrictions – the more options they have to intervene before those homes are lost.
Models that scan property debt, code violations, deferred maintenance and expiring affordability restrictions can flag buildings headed for trouble years before an owner walks away or a subsidy lapses, giving lenders and cities time to step in while the numbers still work.
And the potential extends beyond buildings to helping people remain housed. Los Angeles County built a predictive model with the California Policy Lab to identify residents at high risk of losing their housing and connect them to case managers and flexible cash assistance before a crisis hits. Early results are promising: People who enrolled in the county’s Homelessness Prevention Unit were 71% less likely to enter a homeless shelter or have contact with outreach teams.
That is where AI may be most useful in housing: not replacing human judgment but helping people make better-informed decisions sooner.
Balancing technological gains with community accountability
None of this works without tradeoffs. The data centers powering these tools require significant electricity and water, often in communities already struggling with affordability, and the use of these tools raises important questions about accountability, transparency and human oversight.
As AI becomes more integrated into housing and community development, those considerations need to be a part of how technology is evaluated and deployed. Communities most affected by both the housing crisis and the growth of AI infrastructure deserve a voice in how these tools are built, not only in how they are used.
The housing industry has spent decades doing careful, resource-starved work against a crisis that keeps outrunning it. AI cannot solve the housing crisis. It cannot replace the people, policies, capital or community leadership required to build and preserve affordable homes.
But used responsibly, with real accountability built in from the start, it can help those efforts move faster, stretch limited resources further and give people more time to act before housing opportunities are lost.
Valerie White leads national housing strategic initiatives at LISC, the Local Initiatives Support Corporation.
This column does not necessarily reflect the opinion of HousingWire’s editorial department and its owners. To contact the editor responsible for this piece: [email protected].
