AI and Homelessness Series Part 1: Understanding the Homelessness Application process and AI’s Role in Housing Decisions

Ad/Marketing Communication

This legal article/report forms part of my ongoing legal commentary on the use of artificial intelligence within the justice system. It supports my work in teaching, lecturing, and writing about AI and the law and is published to promote my practice. Not legal advice. Not Direct/Public Access. All instructions via clerks at Doughty Street Chambers. This legal article concerns AI Law.

Introduction

This is the first post in the AI and Homelessness Series. I’ve been thinking about the extent to which Artificial Intelligence (AI) is being used in the homelessness process. Some of my colleagues suggest that local authorities and homeless applicants are starting to adopt AI more broadly, while others argue it’s not being utilised enough. Either way, it seems inevitable that AI will play an increasing role on both sides of the process. In this series, I’ll share my thoughts on the potential legal issues that arise and highlight some practical and legal implications of using AI in homelessness applications.

Whether you’re a legal practitioner, someone working within a local authority, or simply interested in how technology affects crucial decisions impacting vulnerable people, I hope this series will offer valuable insights.

The Homelessness Application Process

Those of us who regularly work in housing law can sometimes take for granted that homelessness applications are widely understood. In truth, the process is highly intricate and demands specialist expertise. This complexity was underscored by Laing LJ in Norton v Haringey London Borough Council [2022] EWCA Civ 1340, where she commented at paragraph 63:

“…This appeal illustrates what is already well known, that housing law can be highly complex….”

I’ve set out a brief overview of the process below to illustrate where AI might potentially be integrated. This is only a simplified snapshot, and there are many variables I’ll explore in future posts.

Initial approach

An individual contacts their local authority for housing assistance because they’re homeless or threatened with homelessness. This may sound straightforward, but it isn’t always. For now, let’s assume the individual has managed to explain their situation to the local authority, and the authority has begun its investigation.

Assessment and Investigation.

The local authority looks into the applicant’s circumstances, including whether they’re legally homeless, eligible for assistance, in priority need (e.g. families with children or vulnerable individuals), and if they’re intentionally homeless.

Initial Decision-Making and Notification.

After assessing the applicant’s situation, the local authority issues a written decision explaining its reasoning and the assistance (if any) it will provide. Depending on the nature of the decision, applicants often have a right to request a review if they disagree.

Review.

Where a review is possible, a more senior officer examines the initial decision to decide whether to uphold it or change it. During this stage, the applicant can provide new evidence or additional representations. The authority may issue a “minded to” decision, meaning it is inclined to decide a certain way, which gives the applicant a chance to submit further information. Fairness is key, and the authority must consider everything with an open mind.

Legal Challenge

If the review isn’t favourable, there may be options to challenge the outcome in the courts on a point of law. If it gets to this stage, legal advice is essential and applicants may be entitled to legal aid to assist in the challenge.

Artificial Intelligence

From even my rough summary above, it’s clear there’s scope for AI to be involved in every stage of the homelessness application process. On the applicant’s side, AI tools might help draft representations or gather evidence. Local authorities might also adopt AI for assessment, initial decision-making, or even carrying out reviews.

I believe there are risks and benefits in both directions. Without careful oversight, vulnerable people could face unintended consequences, such as discrimination or unlawful decisions. At the same time, local authorities could be misled or allocate housing incorrectly, which again impacts those in need. The key question for me is whether AI ought to be used at all, and if so, how can it be harnessed in a way that supports fair and lawful outcomes.