Hey folks, hope you all had a good Memorial Day weekend. We're back with three stories that paint a wider picture than usual — because the AI conversation in real estate isn't just about agents and brokerages anymore. It's about the infrastructure underneath the industry (where Bright MLS's first chief AI officer is calling out a major blind spot), the markets it's reshaping (San Francisco home prices surged 19% year-over-year while 300+ markets face declines — the so-called K-shaped housing market), and the very real backlash it's creating in communities where AI data centers want to be built. Three angles. One bigger picture. Let's dig in.
1. Bright MLS Chief AI Officer: Most Brokerages Are Asking the Wrong Question
Real Estate News published a sharp piece this week pulling quotes from Rajeev Sajja, Bright MLS's first-ever chief AI and product officer, who spoke on a panel at last month's T3 Leadership Summit. His main point cuts through a lot of the noise in the industry: brokerages keep asking "can AI help us?" — and that's the wrong question. The right question, Sajja argues, is: "Are we preparing our data infrastructure to have AI help us at the highest level?"
This is a meaningful reframe. AI tools are only as good as the data they're trained on, and most brokerages — even large ones — are running on fragmented data systems, inconsistent listing entry, dirty CRMs, and disconnected back-office software. According to Sajja, that's the actual ceiling on what AI can deliver, regardless of how many tools you bolt on top. He went further, calling out the industry's culture: "What holds us back today is a legacy system, a legacy structure and legacy technology — and a general avoidance of hard issues." Sajja joined Bright MLS in January after more than two decades at Berkshire Hathaway HomeServices Fox & Roach. His mandate is to make Bright the first "AI-native MLS" — meaning AI isn't bolted on as a feature, it's built into how listings get managed, how insights get surfaced, and how subscribers do their day-to-day work. The MLS already has Bright Solutions launched, with new AI features in the pipeline including tools to simplify listing entry, improve data quality, and surface market insights inside agent workflows.
Why It Matters: This is the structural problem behind everything else we've been covering. When agents say their brokerage's AI tools don't help them as much as ChatGPT does (as last week's Inman Intel survey showed), it's not because the brokerages bought the wrong AI. It's often because the data underneath those tools is messy, incomplete, or trapped in legacy systems that can't speak to each other. Sajja's point applies at every level — from individual agents who haven't cleaned up their CRM in three years, to brokerages running 12 disconnected systems, to MLSs themselves. The agents who win in the next 18 months will be the ones whose data house is in order: clean contact records, structured property notes, organized transaction history, consistent tagging. AI can amplify what you have. It can't fix what isn't there. Spend an afternoon this week auditing your data hygiene. It might be the highest-ROI thing you do all year.
2. The K-Shaped Housing Market Is Here — and AI Is the Dividing Line
The Pinnacle List published a strong spring 2026 market analysis this weekend that crystallizes a pattern we've been seeing build all year: the U.S. housing market is going K-shaped, and AI is the force pulling the top of the K up while the bottom drops further. San Francisco home prices surged 19% year-over-year in March 2026, with a median of $1.7 million — roughly 280% above the national average. The driver? The Bay Area AI boom, which is creating job growth and wealth concentration at a pace the rest of the country can't match.
Meanwhile, Zillow's latest forecast warns that more than 300 U.S. housing markets could see home values decline over the next 12 months. Economists are explicitly calling it "K-shaped" — luxury accelerating at the top, entry-level buyers falling further behind at the bottom. Fannie Mae's May 2026 Housing Forecast reflects the tension: total home sales are projected to rise just 2.1% this year (with a stronger 6.7% bounce in 2027), but new single-family housing starts are forecast to decline 2.4% in 2026, constrained by labor shortages, slow permitting, and elevated material costs. On the construction side, AI platforms are starting to help — auto-generating zoning-compliant site plans and optimizing structural layouts in a fraction of the time it used to take. But the bigger story is policy: on May 20, the U.S. House passed the 21st Century ROAD to Housing Act 396 to 13 — a sweeping bipartisan bill that expands construction loan programs, cuts red tape for builders, and lowers production costs for manufactured homes.
Why It Matters: For agents, this matters because your market is no longer just your market — it's also where it sits on the K. If you're working in an AI-driven metro (Bay Area, Seattle, Austin, parts of NYC), you're in one economy. If you're in one of the 300+ markets where values are softening, you're in a different one. The same playbook doesn't work in both. Buyers in tech-heavy markets are wealthier, more sophisticated, and increasingly willing to pay luxury prices. Buyers in cooling markets are price-sensitive, anxious about job security, and more skeptical. Successful agents in 2026 are going to be the ones who name the dynamic in their market and adjust their pitch accordingly. The ROAD to Housing Act's passage is also worth watching — if it gets through the Senate, it could meaningfully change the construction landscape in the next 12-24 months, especially in markets where labor and permitting have been the chokehold on new supply.
3. 47% of Americans Don't Want an AI Data Center Anywhere Near Them — and the Backlash Is Real
Here's the AI story most brokers haven't focused on yet, and it's one to watch. A Redfin-commissioned Ipsos survey published this month found that 47% of U.S. residents oppose the construction of an AI data center in their neighborhood — making data centers the development Americans least want next door. For comparison, only 37% oppose a new apartment complex and 31% oppose mixed-use development. Even single-family to multi-unit conversions, one of the most contentious housing topics in America, draw just 46% opposition. AI data centers have officially become more unpopular than apartments. A separate Gallup poll cited by industry analysts puts the opposition number even higher at 71%.
The pushback is showing up in real ways. Data Center Watch has identified at least 142 activist groups across 24 states actively working to block data center construction. Numerous projects have been abandoned or postponed. In Virginia — the country's largest data center hub, hosting roughly 13% of all global data centers — Save Prince William County and similar groups have escalated opposition. The pattern is now crossing party lines, with data center opposition helping flip elections in purple-leaning districts across Virginia, Indiana, Ohio, and Pennsylvania. And just last week, hundreds of Vancouver residents marched in protest against new AI data centers announced by Telus, Westbank, and the Canadian government. The complaints from communities are consistent: noise from gas turbines and cooling systems, dramatic increases in electricity and water demand, strained local grids, and lower property values for nearby homes.
Why It Matters: For agents working in markets where data center construction is planned or active, this is a real consideration that affects property values, neighborhood character, and buyer demand. AI data center power demand is projected to rise five-fold over the next decade — reaching 176 gigawatts globally, more than the entire power grid of Australia and the UK combined. That's an enormous build-out, and it has to land somewhere. The communities that say yes will get jobs and tax revenue. The communities that fight back will protect property values but lose economic upside. For listing agents in affected markets, this is a conversation buyers are increasingly asking about — and you need to have an answer. For investors and developers, the policy headwinds around data centers are now real enough to factor into project feasibility. AI's infrastructure is still being built. Where it gets built — and what it costs the surrounding real estate — is going to be one of the most important real estate stories of the next five years.
That's the wrap for this one, folks. The takeaway across all three stories: AI is no longer just a tool sitting on top of real estate. It's reshaping the data plumbing underneath it, redrawing the map of which markets win and lose, and creating very real fights over where the AI infrastructure actually gets built. For agents and brokers, the message is the same as ever — read the room you're working in, know your data, and don't get caught flat-footed by trends that are moving faster than the industry conversation around them. See y'all Friday.