THE LANGUAGE OF DATA SERIES | PART 6 OF 7
Display or Derivative Work: The Only Two Uses
Strip away the acronyms and every use of MLS data falls into one of two buckets. Knowing which bucket you are in tells you most of what you need to know about your rights.
By Saul Klein and John Reilly September 2026
MLS data powers an enormous variety of products: websites, portals, CMAs, dashboards, value estimates, market reports, AI tools. It looks like a zoo. It is actually a coin with two sides, and this article, which stands alone, is about learning to see which side any product is on, because that single judgment determines what rights are required.
Display means showing the data substantially as it arrived: the listing, the photos, the remarks, the price, presented to a human being. Your IDX website is display. Your client portal is display. A portal publishing your listing is display. Display rights are the ones the cooperative shares broadly, because display is how listings get sold, and the rules around display are mostly about attribution, accuracy, currency, and who may not be shown what. Display is governed by the established display programs and their familiar requirements.
Derivative works means creating something new from the data: a comparative market analysis, an automated valuation, a market trend report, a recruiting dashboard, an analytics product, a trained AI model. The listing data goes in; something that is not a listing comes out. Derivative works are where the commercial value concentrates, and they are where the legal exposure concentrates too, because the MLS compilation is a copyrighted database and the right to prepare derivative works from a copyrighted work belongs to the copyright owner unless licensed.
Hold three examples side by side and the principle becomes vivid. Your CMA is a derivative work you are licensed to make; the license rides with your membership and its back office feed. A portal’s instant home value estimate is a derivative work it must be licensed to make; that license is a negotiated commercial contract, not a membership benefit. An AI model trained on listing data is a derivative work at industrial scale, which is exactly why the newest MLS rules address AI by name, and why the question of whether listing data may touch a training environment at all has become the sharpest edge in data licensing.
Here is the practical test for any product, tool, or idea involving MLS data, yours or a vendor’s: is the output a listing being shown, or something new being made? If it is display, look to the display programs and their rules, and your questions will be about attribution, refresh, and disclaimers. If it is a derivative work, look to the license, and your questions will be about scope, retention, and permission. And if it is a derivative work and there is no license, stop. Not slow down. Stop. The entire redesign of MLS data rules now underway, in California, in Texas, and coming to every market, is at bottom an effort to draw this exact line clearly and price each side of it honestly. Learn the two buckets and you can read every rule that follows.
This article stands on its own and is also part of the practitioner series “The Language of Data”. Copyright 2026 Saul Klein and John Reilly. All rights reserved. Commentary and general information, not legal advice.
