Overview

Using a Colorado housing application, this article traces a Covered ADMT that sets a $2,500 security deposit as the sole basis, compared with $1,500–$2,500 paid by other consumers. It tests coverage, developer-integrator-deployer responsibility, versioned evidence, and a specific explanation, while stressing that the rulemaking example does not establish the decision's lawfulness.

Learning outcomes

In the housing example, the score materially influences a consequential term because it is the sole basis for the $2,500 deposit.
A useful explanation identifies the decision, system role, sole-basis reliance, principal reason, and relevant score or inference.
End-to-end accountability fails when coverage, handoffs, versions, inputs, or consumer-facing evidence cannot be connected.

Keywords

ColoradoADMTHousingSecurity depositEvidenceExplanation

Checkpoint: Can You Follow One Decision End to End?

Colorado's automated-decision rules become easier to judge when the legal chain is tested against one ordinary event: a housing provider sets a security deposit with the help of a computed score. The useful question is not whether the provider has an AI policy, but whether a reader can follow the applicant's data, the system's output, the decision's effect, the organizations involved, and the evidence that would later support an explanation. The housing scenario below connects those established ideas and shows where a consumer-facing remedy must begin.

A housing deposit decision, traced from input to consequence

Assume a rental housing provider receives an application for a three-bedroom apartment. It uses a Covered ADMT to analyze information supplied by the applicant and sets the security deposit at $2,500. Other consumers seeking the same size unit have paid between $1,500 and $2,500. The system's stated reason is that the applicant lacks recent rental history, and the provider uses the system as the sole basis for the amount.

The Department of Law's rulemaking document uses a materially similar housing example to illustrate the content of an adverse-outcome disclosure. The example is not a finding that this way of setting a deposit is lawful. Its value is analytical: it lets us test the connection among personal data, computation, output, decision, and effect without inventing a technical architecture that the sources do not describe.

The first question is whether the technology is ADMT. The system analyzes information about an individual and computes a result that guides the deposit decision. It is not merely storing the application or arranging documents for a human to read. Its output affects the provider's judgment about a material term of housing. That satisfies the functional starting point of the statutory framework.

The second question is material influence. The output is not an incidental signal that the provider ignores. It is the sole basis for setting the amount. The connection to the outcome is therefore direct: the computed result changes the price-like term offered to the applicant. Even if an employee clicks the final approval button, the output has done more than assist with clerical processing.

The third question is consequentiality. Residential real estate is a covered domain, and the statute includes decisions about access, eligibility, selection, and material terms that may materially limit or fundamentally alter that access. A higher security deposit can matter even when the applicant is not rejected outright. For a household with limited cash, the added amount may delay or effectively prevent the lease. The legal significance comes from the effect on access to housing, not from the prestige or complexity of the software.

The final question is whether an exclusion removes the use from Part 17. Nothing in this fact pattern suggests a routine scheduling function, a document-organizing tool, fraud-prevention activity, or another listed exclusion. The system is being used to set a material housing term.

Published by DeepMox

FAQ

Why can a higher housing deposit be a consequential term under Colorado's ADMT Act?

A higher housing deposit can be consequential because a differentiated material term may materially limit, delay, effectively deny, or fundamentally alter access to housing.

What facts make the housing example a Covered ADMT use?

In the example, applicant data produces a computed result that is the sole basis for a $2,500 deposit in a residential-housing decision, and no supplied exclusion applies.

What should a Colorado housing explanation say about an AI-set deposit?

It should identify the deposit decision, the system's role and sole-basis use, and the specific reason—insufficient recent rental history—along with any relevant score or inference.

Previous9 minColorado AI Decision Records

Next9 minColorado ADMT Consumer Remedies

Colorado ADMT Housing Deposit

Checkpoint: Can You Follow One Decision End to End?

A housing deposit decision, traced from input to consequence

FAQ