Colorado SB26-189 for Leasing Agents Using AI (2027)

The short answer, as a checklist

Starting January 1, 2027, Colorado's SB26-189 applies to AI and other automated tools that materially influence a decision about leasing residential real estate in Colorado. If a tool scores, ranks, classifies, or recommends renters in a way that shapes who gets a unit, plan to treat it as subject to the law. If a tool only schedules, routes, organizes, summarizes, translates, or drafts for a person to review, the statute carves much of that out.

This guide is for leasing agents and small property managers who rent in Colorado or are thinking about adopting AI tools there. It was checked in October 2026 against the signed law and the Attorney General's published materials. It is general information, not legal advice, and the Attorney General's rules are still in draft.

Work through this list before the end of the year.

• List every tool that touches a renter's information, from your website chat to your spreadsheets.
• For each tool, write down what it outputs: a score, a rank, a pass or fail flag, a recommendation, or just organized information.
• Mark any tool whose output affects who gets approved, denied, or offered different terms as likely subject to the law.
• For tools the law applies to, plan a clear notice where renters interact with you, and a way to send an adverse-outcome notice within 30 days.
• Name the person who can review and change a decision when a renter asks, and make sure they don't just follow the tool's output.
• Plan to keep records of each covered decision for at least three years.
• Ask each vendor for the documentation the law requires them to give you, and keep it.
• Ask your attorney to confirm which of your tools fall inside the law.

Who and what the law covers

SB26-189 was signed on May 14, 2026. Its duties apply to consequential decisions made on or after January 1, 2027. A consequential decision includes one that relates to a consumer's access to, eligibility for, or selection for a covered domain, and the lease or purchase of residential real estate in Colorado is one of the covered domains. A renter whose access to housing in Colorado is being evaluated counts as a consumer.

The law talks about automated decision-making technology, or ADMT: technology that processes personal data and produces predictions, recommendations, classifications, rankings, scores, or other output that is used to make, guide, or assist a decision about an individual. Your duties attach when that technology is used to materially influence a consequential decision, which the law calls a covered ADMT.

Materially influence has two parts. The tool's output has to be a non-trivial factor in the decision, and it has to affect the outcome, for example by constraining, ranking, scoring, recommending, or classifying. Incidental, trivial, or clerical uses are excluded. We did not find a size-based exemption for small landlords in the text of the act, so assume a one-person leasing business is not automatically outside it and ask your attorney.

What the law carves out

The carve-outs matter for leasing agents because much everyday AI help is administrative. Whether a tool lands inside or outside the law depends on what it produces and how you use it, not on whether it uses AI.

The Attorney General's October 6, 2026 interim draft rules give two lease examples. A tool that sorts applicants into positive and negative groups would be covered. A spreadsheet that organizes applicants by the unit size they asked for and the date they applied, with no scoring, would not count as automated decision-making technology. These are draft examples, not final rules.

• A tool used solely to summarize, organize, translate, draft, route, or present information for human review or administrative processing is not automated decision-making technology under the law.
• Routine scheduling, administrative routing, customer service triage, communicating decisions, and workflow management are not consequential decisions when they don't materially influence who gets housing.
• Using automated decision-making technology to summarize, organize, or present information for human review is excluded when the system doesn't produce a score, ranking, recommendation, classification, prediction, or other inference that materially influences the outcome.
• Spreadsheets that need human analysis and don't use machine learning, foundation models, or large language models are excluded.
• General chat technology that gives information, answers questions, or makes referrals is excluded only if it is not configured or marketed for consequential decisions and an acceptable use policy bars using its output in one.

What you must do if the law applies to a tool

If you use automated decision-making technology to materially influence a lease decision, the law calls you a deployer and puts several duties on you. The vendor, which the law calls a developer, has its own documentation duties to you.

An adverse outcome includes denying a renter, and also a decision that results in materially less favorable terms than similarly situated renters get. Whether a given outcome counts depends on the facts, so ask your attorney about edge cases such as a higher deposit.

• Notice before use: tell renters, clearly and conspicuously, that you use or will use automated decision-making technology in a decision affecting them. A prominent public notice at the places renters interact with you, such as a link next to the application or chat, satisfies this.
• Notice after an adverse outcome: within 30 days, give a plain-language description of the decision and the tool's role, how to request more information about the tool and its inputs, and an explanation of the renter's rights.
• Correction and review on request: after an adverse outcome, give instructions for getting and correcting inaccurate personal data, and an opportunity for meaningful human review and reconsideration, to the extent commercially reasonable.
• Meaningful human review means a person you designate who can approve, modify, or override the decision, looks at the primary evidence, is trained, doesn't default to the system's output, and has enough information to understand what the output is for and its limits.
• Records: keep records of covered decisions for at least three years.

Enforcement, timing, and what is still unsettled

Only the Attorney General enforces the law. A violation is a deceptive trade practice, and the act creates no private right of action. Before an enforcement action the Attorney General issues a notice of violation when a cure is possible, with 60 days to fix it. No cure period is required for knowing or repeated violations, and the cure provision is repealed on January 1, 2030.

The act does not replace other law. Using an automated tool is not a defense to discrimination or consumer protection claims, and following SB26-189 does not excuse non-compliance with fair housing rules. A leasing agent stays responsible for the decision.

The Attorney General has proposed rules on what counts as materially influencing a decision and on the content of adverse-outcome notices. The interim draft is dated October 6, 2026, with comments open through October 26, 2026, and rules due by January 1, 2027. One draft rule would presume a tool materially influences a decision when its output constrains options, applies a threshold, ranks, scores, classifies, or recommends, and the decision maker reviews it and decides consistently with it. Expect details to change before the rules are final, and check the Attorney General's AI page and the Colorado Division of Real Estate summary before you finalize your process.

How Rentalot approaches this

Rentalot's public profile and structured pre-screening collect a renter's answers and organize them for you. You decide who gets a showing, an application, or a lease. Rentalot does not approve, reject, or rank renters.

Some pre-screening questions let you set minimums, such as an income multiple or a pet limit. If you use a minimum like that to sort renters, treat it as a decision rule: put it on your tool list, decide whether it falls inside the law, and have your attorney confirm. This post describes the statute and the Attorney General's draft, not a legal opinion about any product.

Related on Rentalot: AI prospect pre-screening · Fair housing compliance for AI tools · AI tenant screening guide