Don’t miss: Inside AI Episode 5: Sam DeBord, CEO of RESO, explains why AI needs a common language for real estate

For years, I’ve heard people in real estate complain about siloed technology. Tools that agents use don’t talk to each other making repetitive entry of data the standard. Along came RESO, the Real Estate Standards Organization, and things began to change. The data standards they created enable technology to finally speak the same language.

Now AI comes along, and suddenly there’s a new version of that argument: If artificial intelligence can understand almost anything, why do we still need real estate data standards? It sounds logical, but after my conversation with Sam DeBord, CEO of the Real Estate Standards Organization, better known as RESO, I’m convinced the opposite may be true.

AI may make standards more important than ever.

That was one of my biggest takeaways from Episode 5 of Inside AI with The REAL AI Guy. Sam has a great way of explaining standards. Think Bluetooth or Wi-Fi. They’re common protocols that allow different technologies to connect and work together. RESO does something similar for real estate data.

Today, RESO has member organizations with a footprint that reaches about 30 countries. Its standards give real estate technologies a common language, even when the systems, companies and markets using that data may be very different.

AI still needs a Rosetta Stone

There’s a tendency to think of AI as all-knowing. Feed it enough information and eventually it will figure everything out. Sam pointed out the problem with that assumption.

The real estate industry has roughly 500 MLS organizations, thousands of standardized data fields and millions of additional local data elements. RESO has processed about 2 million data elements through certification, representing roughly 20 million data points.

Could AI try to interpret all of that independently? Sure. But why would we want it to reinvent something the industry has already spent years organizing?

Sam described standards as a kind of “Rosetta Stone” that gives technology a foundation for understanding real estate data. AI can build on top of that foundation rather than trying to create its own definition for everything.

Then he made an even better point. “If OpenAI and Google and Microsoft and Anthropic all decided to create their own standards, what would we have?” Sam asked.

We wouldn’t have a standard. That’s the issue.

The future of real estate won’t run on one AI company, one large language model or one technology ecosystem. Different organizations will use ChatGPT, Claude, Gemini and whatever comes next, which means those systems still need a common language.

The data isn’t free just because AI can find it

Here’s where the conversation became particularly important for brokers, MLS leaders and anyone responsible for real estate data. AI accelerates everything, including the problems we already have.

MLS data has always been licensed for specific uses by specific users. But AI systems are very good at consuming enormous amounts of information, often without the end user understanding where that information came from or what rights were attached to it.

Sam described AI as being known for “slurping up all the data in the world.” That creates a major problem for real estate.

RESO is working with the industry on ways to embed governance information directly into the data feed itself. Instead of licensing rules living in a PDF buried on a website or attached to an email, an AI system could receive instructions within the data explaining who may use it and what the data is licensed to do.

That doesn’t magically solve every legal or technical issue, but it creates something AI desperately needs: context that creates a guardrail. The system can’t operate as well if there are no rules attached to distributed information.

Agents need to understand this too. Just because you can copy information into an AI tool doesn’t mean you should, especially when you don’t know how that information is protected, stored or used.

This is why I keep coming back to one of the most important things I teach about AI: The best AI tools are also the safest AI tools. They are also the most transparent ones.

RESO is giving AI a real estate brain

One of the more fascinating parts of my conversation with Sam was learning what RESO is already building. I knew the organization has created software development kits, certification tools, and reference servers. I did not know they also have an AI MCP server.

MCP, or Model Context Protocol, is becoming increasingly important because it provides a structured way for AI systems to connect with outside tools and data. RESO’s approach is especially interesting because the AI isn’t simply being pointed at a database and told to figure it out.

The standards are built into the foundation.

Sam explained that a data user could connect an AI system such as ChatGPT, Claude or Gemini and begin asking questions about a real estate marketplace. The RESO MCP server already understands how the data is structured and how the question should be asked.

That’s a huge distinction. A random AI trying to navigate an unfamiliar API may eventually find the answer, but an AI that already understands the rules starts from a much better place.

AI with 1,500 rules

The other part of RESO’s work that jumped out at me was how much effort went into teaching its AI how RESO thinks. Sam said RESO CTO Josh Darnell spent significant time “mentoring” its “Super Claude” on RESO specifications, testing systems, organizational practices and constraints.

RESO then turned those principles into persistent memory and policies for the AI to follow. The organization doesn’t declare one competitor better than another. It doesn’t say MLSs are more important than brokers. It has very specific ways of describing products, organizations and standards.

Eventually, Sam said, that library grew to around 1,500 policies and rules.

That may be one of the most useful lessons from this entire interview because the future of business AI isn’t simply giving employees access to ChatGPT. It’s teaching AI how your organization operates.

What are your rules? What are your policies? How do you make decisions? What information can the AI use, and where does a human need to review the work?

The companies that answer those questions are going to get far more value from AI than companies that simply hand everybody a chatbot.

Months of work in two days

Then Sam gave me one of those examples that makes you realize how quickly this technology is changing what small organizations can accomplish.

RESO had an older Data Dictionary Wiki containing thousands of interconnected data elements. Rebuilding something like that traditionally would have been a significant project, but Sam said Josh rebuilt the entire site in less than two days.

And this wasn’t simply a rough mockup. The process included multiple iterations, improvements, styling changes and user-interface adjustments. A project that might once have taken months was compressed into days.

RESO also built what had been envisioned as a two-year technical roadmap in roughly four or five months, with significantly more included than originally planned.

That’s extraordinary, but Sam immediately added the caveat that matters most. AI still makes mistakes, including potentially critical ones, so RESO keeps experienced engineers in the loop – a human in the middle – to review what AI creates.

That’s exactly the balance I keep stressing. Use AI aggressively to save time, but don’t sacrifice accuracy for speed.

Please don’t vibe code the infrastructure

We eventually got into agentic AI, AI systems that don’t simply answer questions but can take actions on your behalf. This is where I admitted something to Sam: Agentic AI scares the bejeebies out of me.

The opportunity is enormous, but so is the risk when we give software permission to act autonomously. It’s also why I keep warning real estate professionals about vibe coding business-critical systems simply because AI now makes it possible.

Sam agreed. “Nothing we build is vibe coding,” he said.

Building a fun personal app is one thing. Building infrastructure that other businesses will rely on is completely different. Sam warned against ending up with “a bunch of spaghetti code” that won’t work reliably with the next piece of technology someone builds.

Just because AI lets you build something doesn’t mean you suddenly became a software engineer. That distinction is going to become increasingly important as these tools become more powerful.

Your company needs a second brain

One of my favorite parts of the conversation had nothing to do with data standards. Sam talked about capturing institutional knowledge.

Every organization has employees who know things nobody ever wrote down. They know why a process works a certain way, who needs to be called when something goes wrong and all the little nuances that never make it into the employee manual.

RESO is experimenting with using AI to capture that knowledge. Instead of forcing employees to create massive documentation, imagine spending 30 minutes a week explaining to AI what you did, why you did it and the process you undertook.

Over time, you build a searchable organizational memory. If someone takes a vacation, changes jobs or retires, the knowledge doesn’t disappear with them.

My WAV Group partner Victor Lund has written about a similar concept for company founders: create your second brain or digital twin. Teach an AI what you know, how you think and how you make decisions, so the next generation of leadership doesn’t have to start from scratch.

Real estate has lost an enormous amount of institutional knowledge over the decades because there was never an easy way to preserve it. AI changes that.

Standards may matter more than ever

Toward the end of our conversation, I asked Sam to look two or three years ahead. His answer pointed to the early days of the web, when information that had been locked behind geography, technology and cost suddenly became accessible to almost anyone.

Sam thinks AI could create another leap like that, especially in housing markets where property information remains fragmented across disconnected systems. RESO won’t build all those future applications because that isn’t its role.

Its job is to keep building the foundation: standards, APIs, software development kits, and even governance frameworks and tools that allow everybody else to innovate on top of reliable real estate data.

And that brings us right back to the question we started with. If AI can understand almost anything, do we still need standards?

Absolutely.

The smarter AI gets, the more systems it connects with, the more data it consumes and the more actions it can take. That makes it even more important that those systems understand the same language, the same rules and the same meaning behind the data.

AI isn’t making RESO standards obsolete. It may finally be showing us why we needed them all along.

Watch Episode 5 of Inside AI with The REAL AI Guy to hear my full conversation with Sam DeBord, CEO of RESO.

# # #

Kevin Hawkins, a partner with WAV Group, is editor and co-founder of REAL AI, real estate’s No. 1 AI newsletter, its weekly podcast and the Inside AI with The REAL AI Guy video interview series. He has written more about AI than anyone in real estate. Hawkins is the Amazon bestselling author of The REAL AI Guide for Real Estate Agents and is known throughout the industry as The REAL AI Guy, teaching thousands of real estate agents how to use AI in the best and safest ways.