This week’s Inman Tech Roundup focuses on artificial intelligence getting closer to real estate agents’ work — from CRMs that talk to outside AI assistants to tools that turn an address into a client-ready property briefing. But a new report on AI-edited listing photos reminds us that the same technology is outpacing the rules meant to keep it honest.
Here’s what you need to know:
Lofty joins the MCP rush
Lofty, the Phoenix-based brokerage platform formerly known as Chime, launched Lofty MCP Server on Monday. The server is a connector that lets brokerages securely hook their own AI agents and internal systems into the Lofty platform.
The tool is built on the Model Context Protocol, the open standard Anthropic introduced in late 2024 that’s quickly become a common way to connect AI models to outside software. Lofty says it requires no custom integration code and lets MCP-compatible AI clients read Lofty data, analyze activity and, where authorized, take action inside the platform.
In one example the company offered, a broker could ask an in-house agent to flag leads that showed strong buying intent in the past 30 days but haven’t been contacted in a week.
Joe Chen
The server works alongside Lofty AOS, the company’s own agentic AI system, rather than replacing it. The pitch is that Lofty stays the system of record even as brokerages build or buy AI agents elsewhere.
“Brokerage leaders understand the transformational power of AI, but AI is only as valuable as the business context it can access,” Joe Chen, Lofty’s CEO, said in a statement. “Lofty brings together a brokerage’s people, data and workflows into one operating system, and now we’re making that intelligence securely accessible to the AI agents our customers choose to use.”
Lofty isn’t alone. MCP servers have quickly become table stakes for proptech in 2026, with Rechat and Inside Real Estate among the latest to roll one out.
Data providers got there first. ATTOM launched its MCP server in January, PriceHubble built one for property data and valuations, and Local Logic followed earlier this month.
What this means for agents
For real estate agents, the practical upshot is that they may soon be able to query and update their CRM from whatever AI assistant they already use, like Claude or ChatGPT, instead of toggling between tools. How much of that access reaches individual agents, though, will depend on what their brokerage chooses to connect and authorize.
Author’s note: In August, Lofty was named Inman’s official AI partner.
Land id turns an address into an AI property briefing
Land id, the mapping and land analysis platform, last week released The Land id, an AI product that assembles what’s known about a property and explains what matters in plain language.
Starting from an address, the tool pulls public records, licensed data and users’ own content, then surfaces insights on the map. Land id pitches it as filling gaps that traditional CMAs, memoranda and investor decks often miss. It also generates portal-ready property descriptions and shareable visualizations that work on phones, tablets and browsers.
The product runs on Land id Intelligence, the company’s proprietary AI engine, and targets everyone from residential agents to land and ranch brokerages, multifamily operators and investors.
Steve Roberson
“This industry has never had a shortage of property data. It has had a shortage of ways to use it. AI is what finally closed that gap,” said Steve Roberson, founder of Land id. “We have spent years assembling the right data and building the intelligence on top of it, and The Land id is the result. One tool, one record and answers where there used to be a stack of subscriptions that never talked to each other.”
The Land id is available now to Pro and Premium subscribers, with Basic users getting access for a limited time. The company says it has more than 50,000 paid users.
What this means for agents
Land id’s new tool promises to turn a pile of property data into talking points and visuals they can put in front of clients in minutes. This could sharpen listing presentations and buyer consultations.
CubiCasa and Restb.ai team up to speed listing prep
CubiCasa and Restb.ai on Monday announced free rush processing of CubiCasa floor plans for MLSs that use both companies’ technology.
The perk cuts delivery time from 24 hours to under 12, and the companies estimate it covers about 500,000 agents and brokers.
Thirty MLSs qualify initially. To be eligible, an MLS must be in CubiCasa’s MLS Partnership Program and subscribe to Restb.ai’s AI photo tagging within its main listing input platform.
Jeff Allen
The two companies handle different pieces of the listing workflow. CubiCasa turns smartphone scans into floor plans and property tours, while Restb.ai uses computer vision to convert listing photos into structured data for search, compliance and data quality.
Restb.ai says its MLS partnerships reach more than 1 million agents across the U.S. and Canada.
“Speed matters when an agent is getting a listing ready for market,” said Jeff Allen, president of CubiCasa. “Giving agents priority processing at no additional cost means they can receive their floor plans in less than 12 hours.”
What this means for agents
For agents in the 30 eligible MLSs, the practical change is getting floor plans back overnight instead of in a day. This could shave time off listing prep when a seller wants to hit the market fast.
The catch is that eligibility depends entirely on their MLS’s vendor choices, so agents elsewhere won’t see any change unless their MLS signs on with both companies.
AI-edited listing photos are outpacing disclosure rules
More than 90 percent of AI-altered listing photos on major U.S. portals don’t disclose to buyers or renters that AI was used, according to new research from Coraly.ai.
The AI proptech analyzed nearly 40,000 primary listing photos from Zillow, Redfin, Realtor.com and Homes.com in the first quarter of 2026, the same quarter California’s law requiring disclosure of AI-altered listing photos took effect.
Coraly found that about 11 percent of listings used AI-edited primary photos.
Rates ranged from 8.7 percent on Realtor.com to 12.4 percent on Homes.com, with Zillow at 11 percent and Redfin at 11.2 percent. Sky replacement was by far the most common edit, appearing in 69 percent of altered images.
The findings land as scrutiny of AI-enhanced listings grows. New York’s Department of State issued a consumer alert on AI-generated listings last November, and New York City Mayor Zohran Mamdani proposed disclosure rules for rental listings in July.
Coraly connects marketplace benchmarking with practical real-estate AI for property portals, MLSs and real-estate organizations. The company has run the same study in the U.K., Germany and other markets.
What this means for agents
For agents, the findings are a warning: The quick sky alteration or virtual staging job they’ve come to treat as routine could put them on the wrong side of disclosure rules, especially in California, where the law now requires labeling AI-altered photos.
With more states and cities eyeing similar rules, adding a clear “digitally altered” label now is a cheap hedge against both regulatory trouble and the buyer backlash that comes with “housefished” listings.
