Showing AI + Community + Homelessness.

How AI + Community and Homelessness Intersect

Picture a person who needs a shelter bed tonight and is told to start with a chatbot. AI is entering the places people depend on most: shelter intake, benefit help, housing applications and social services. AI-driven tools can speed up a connection to help. They can also shut people out when they assume a working phone, steady internet and comfort with a screen. The same pressures that cost people their homes also cut them off from those tools.

AI + Communityhow AI changes civic lifeHomelessnessthe most visible housing crisis
AI + Community and Homelessness: Decisions people can't see or challenge
Where they meet: where daily life compounds
  1. AI + Community and Homelessness: Decisions people can't see or challenge
01

AI + Community + Homelessness

The first contact is a chatbot

Agencies and nonprofits increasingly use AI-driven intake tools to handle demand. Using them takes a chain of four things: a device, a connection, digital comfort and language access. Housing instability breaks that chain. A person without a phone or a mailing address may be screened out before a human ever sees the case. For people here, the front door to help can now be a screen they cannot reach.

02

AI + Community + Homelessness

Old patterns, repeated faster

AI trained on past data can repeat past disparities. That can change who is sent to services, who is flagged for review and who is put at the back of the line. People experiencing homelessness already move through programs shaped by race, income and earlier contact with public agencies. In Mecklenburg County in 2025, 74% of those experiencing homelessness were Black, according to Mecklenburg County. Automated scoring can quietly add to those gaps, with no clear way to appeal.

03

AI + Community + Homelessness

Decisions people can't see or challenge

Suppose a risk score or an automated form decides who gets a shelter referral or a housing voucher. That decision can be harder to question than one made by a person. Responsible use of AI asks for three things. People need a human option. They need clear rules for how decisions are made. And they need a plain way to understand and contest the result. Without these, people in crisis have no one to ask why.

How automation can deepen housing instability

  1. 01

    The housing crisis starts

    Rent costs, a lost job or a family breakup push someone toward losing housing, and phones and internet slip away.

  2. 02

    Digital barriers show up

    The agency moves its intake to AI-driven tools, and the person cannot get online to use them.

  3. 03

    Automated screening filters them out

    An AI tool flags an unfinished form or an odd address and sends the person away without human review.

  4. 04

    The gap widens and help is delayed

    Each failed try costs time and trust, and the person may stop asking and slide deeper into crisis.

Designing AI service tools with a human option, offline paths and a real review of unresolved cases can break this cycle at the intake step.

By the numbers

  • 7,0007,000 job postings requiring AI skills in Charlotte metro area in 2025
  • 2,0182,018 people in Charlotte-Mecklenburg identified as experiencing homelessness per the 2026 Point-in-Time Count
  • 44%44% of CMS staff said they’ve used AI at work, while 59% also expressed concern about academic integrity and its impact on interpersonal skills
  • 74%74% of those experiencing homelessness in Mecklenburg County in 2025 were Black

Folio links, articles, and reports

Sources & methodology15 referenced sources

Based on fftc:FWD Folio material for these topics and public data, with broader context on how AI service design and housing instability connect.

AI assists with organizing and explaining selected fftc:FWD source material. It does not replace review by FFTC staff or the judgment of residents and civic partners.