A major brokerage CEO said the quiet part out loud this week: he expects AI to mean fewer agents. Meanwhile Zillow is building a home for your buyers that you don't own. Big week. Let's make sure you're on the right side of it.
Here's what you need to know this week:
Elliman goes all-in on AI, expects layoffs -- the CEO says the industry needs fewer agents
Zillow's new buyer hub wants to run the transaction -- with a slot reserved for you
82% of agents use AI, only 17% see real impact -- the adoption gap is now the whole story
Your clients are pricing homes with ChatGPT -- and walking in with the wrong numbers
Let's get into it.
1. Elliman Bets the Brokerage on AI
TL;DR: Douglas Elliman announced a company-wide AI transformation built on Google Cloud, launched a new intelligence subsidiary called Elius, and its CEO said flatly that AI will reduce the need for so many agents.
Elius, announced July 8, will turn Elliman's decades of proprietary deal data into AI products and market intelligence sold to outside customers, not just used in-house.
CEO Michael Liebowitz told Bloomberg the tech upgrade will lead to layoffs at the firm, with the exact number decided after implementation.
The company expects "significant savings in non-commission operating expenses" over the next three years through automation and workflow consolidation.
Liebowitz went further than any major brokerage CEO to date, saying he expects AI to shrink the number of agents the industry needs (The Real Deal).
Our take: Ignore the panic headlines and read what actually happened: a brokerage decided its data is worth more than its headcount. The agents who survive that math are the ones AI can't substitute for, the ones with relationships, negotiation skill, and local judgment. If your daily output is tasks a workflow can automate, this announcement was about you. If it's advice and advocacy, you just got more valuable.
2. Zillow Builds a Home for Your Buyer
TL;DR: Zillow's new hub walks buyers from first search to closing inside the Zillow app, with your role reduced to a contact card in their "shopper's team."
The hub organizes the purchase into four milestones (budget, search, offer, closing) and auto-updates as buyers complete steps. Live now on iOS and Android, web version coming.
It centralizes documents, tasks, local market data, and a one-year price forecast, plus a Verified Pre-approval tied to listings.
The "shopper's team" feature slots in the buyer's agent and lender, and connects unrepresented buyers to agents through Zillow's Agent Finder (Inman).
Our take: Zillow is making a play to be the operating system of the transaction, which makes you a feature inside their product. The counter isn't boycotting the app your buyers already use. It's making sure the guidance layer stays yours: your own milestone checklist, your own market updates, your own closing roadmap delivered before Zillow's notifications do it for you. The agent who explains what happens next is the one clients remember at referral time.
3. The AI Adoption Gap Is the Real Story
TL;DR: Nearly every agent now "uses AI," but the business results are concentrating in a small group that uses it differently than everyone else.
82% of agents have integrated AI into their business as of 2026, per NAR subsidiary RPR (HousingWire). But NAR's Technology Survey found only 17% report a significant positive impact, while 46% see no noticeable difference.
The gap is closing for power users: agents telling Inman Intel that AI made them "significantly more productive" jumped from 11% in December 2024 to 28% as of April 2026 (Inman).
Time savings are real but unevenly claimed: 68% of agents save at least an hour a week with AI, and 34% save more than four hours (RPR, 2026).
The dividing line isn't which tool you bought. It's whether AI is doing occasional writing tasks or running repeatable systems (lead follow-up, CMA prep, content pipelines).
Our take: "Using AI" stopped being a differentiator the moment 8 in 10 agents could say it. The 17% getting real results aren't better prompters, they're systematizers: they picked two or three workflows that touch revenue and automated them end to end. Pick one this week. Listing description polish doesn't win listings. A follow-up system that touches every lead within five minutes does.
4. Your Clients Are Pricing Homes with ChatGPT
TL;DR: Brokers report a surge of buyers and sellers arriving with AI-generated price opinions that are confidently wrong, and it's complicating negotiations.
Agents told CNBC that clients increasingly use AI chatbots to decide what a home is worth and what to offer, then anchor hard to those numbers.
The estimates are often stale or wrong because general-purpose chatbots lack current MLS data, recent comps, and hyperlocal context.
The problem cuts both ways: sellers overpricing off AI optimism, buyers lowballing off AI pessimism, and agents stuck re-educating both.
Our take: Don't fight the chatbot, out-data it. When a client quotes an AI price, ask what comps it used. It usually can't say, and that's your opening. Walk them through three real closed comps from the last 60 days and the conversation resets. This is also an argument for sending clients your own market data regularly, because the agent who informs them first sets the anchor.
Get the Full System
This week's theme is the gap between using AI and getting results from it. The playbook exists to close exactly that gap, with automation blueprints and a 30-day implementation plan that turn one-off prompts into systems that run your follow-up, content, and CMA prep.
Get it for $49: AI Realtor Edge Playbook on Gumroad
Quick Hits
Compass rolls out its AI platform across brands -- The AI-powered Home Platform (CMA, pricing presentations, marketing center, AI coach) is now live for agents at Coldwell Banker Realty, Corcoran, Sotheby's International Realty, and @properties, six months after the Anywhere merger closed. (Real Estate News)
Shilo launches AI 1:1 coaching for agents -- The tool builds an agenda from your call history, runs a live voice-to-voice coaching session, and remembers it for next time. No manager required. (Inman)
Realty ONE Group debuts ZONE Pro -- A proprietary AI operating system bundling a growth coach, referral network, and business automation for its 20,000+ agents. (Inman)
An AI mortgage brokerage just raised a seed round -- Y Combinator-backed Ralo launched with $2.9M, using AI loan officers for rate shopping and pre-approvals, claiming average closings around 15 days. (Real Estate News)
Prompt of the Week
For the next client who walks in with a ChatGPT price opinion. Paste your real comp data and let AI build the counter-narrative for you:
"You are a real estate pricing analyst. Here are 3-5 closed comps from the last 60 days in [NEIGHBORHOOD]: [paste address, sale price, days on market, and key features for each]. Write a plain-English pricing explanation for a [buyer/seller] who believes this home is worth [CLIENT'S NUMBER] based on an AI chatbot estimate. Respectfully explain what the actual comps show, why chatbot estimates miss [current inventory/recent sales/condition adjustments], and end with a suggested price range. Keep it under 250 words and use a confident, friendly tone."
It works because you're not arguing with their AI, you're replacing its missing data with real comps and letting the numbers do the correcting.
Market Pulse
Rates: 30-year fixed at 6.49% as of July 9, up from 6.43% the prior week (Freddie Mac). A year ago: 6.72%. Sales: June existing-home sales at a 4.09M SAAR, down 2.4% from May but up 2.8% year-over-year (NAR). Inventory: 1.56M units, a 4.6-month supply, essentially flat from May. Prices: Median existing-home price $440,600, up 1.8% YoY. That's 36 straight months of annual increases. Bottom line: A sideways market: more supply than the frenzy years, prices still grinding up, rates stuck in the mid-6s. Deals go to agents who set realistic expectations early, which is exactly what stories 3 and 4 above are about.
Which story hits closest to home this week: Elliman's bet, Zillow's hub, or the ChatGPT pricing problem? Hit reply and tell us.
Stay ahead, AI Realtor Edge www.AIRealtorEdge.com
Disclaimer: AI Realtor Edge is for informational and educational purposes only. It is not legal, financial, tax, or real estate advice. Results are not guaranteed and will vary. Always consult licensed professionals for your specific situation.