5 Things to Know About AI This Week
A quick scan of what happened in AI this week that's worth knowing if you work in real estate.
1. Anthropic Releases Claude Opus 4.7 and a New Design Tool
Anthropic launched Claude Opus 4.7 on Thursday, calling it the company's most capable publicly available model. Alongside it, Anthropic shipped Claude Design, a new product that turns conversational prompts into polished slide decks, one-pagers, and marketing collateral. For real estate teams, that second announcement may matter more than the model update itself: it puts an AI-powered design tool directly into the same chat window many agents are already using to draft listing copy and client emails.
2. OpenAI Pivots Hard Toward Business Users
In interviews this week, OpenAI executives said business customers now make up 40% of revenue, up from 20% in 2024, and that number is expected to hit 50% by year-end. The company also previewed a new model codenamed Spud built for "high-value professional work." Translation for real estate: the consumer-facing ChatGPT most agents use is being rebuilt around business workflows, which should mean better tools for things like email drafting, document review, and transaction support in the months ahead.
3. Houston's MLS Opens Real-Time Data to AI Platforms
HAR.com announced a partnership with Repliers this week that makes the real estate data company the exclusive platform for on-demand MLS data delivery. HAR is also adding some of its own proprietary data, including page views, lead stats, and showing activity, to the shared API. It's another signal that MLSs are opening up their data so AI tools can pull from live listings instead of stale ones. Expect more MLSs to follow, and expect the gap between AI-ready brokerages and everyone else to keep widening.
4. FBI Reports $275 Million in Real Estate Fraud, with AI Cited as a Driver
The FBI's Internet Crime Complaint Center released its 2025 numbers on April 13, showing more than $275 million in real estate fraud losses across 12,368 complaints, both figures up significantly from 2024. The report specifically calls out AI as a reason fraud is harder to detect, noting that synthetic profiles, deepfake videos, and AI-generated conversations are becoming convincing enough to fool seasoned professionals. Every agent should treat this as a reminder to verify wire instructions by phone on a known number, and to build identity verification into every deal, especially for out-of-state sellers, vacant land, and remote closings.
5. Stanford Says the U.S.-China AI Gap Has Nearly Closed
Stanford's Human-Centered AI Institute released its 2026 AI Index report this week, showing the performance gap between top U.S. and Chinese AI models has shrunk to just 2.7%, down from double-digit leads only a few years ago. For real estate professionals, the bigger-picture takeaway is that the AI tools market is getting more competitive globally, which usually means faster innovation, more options, and downward pressure on pricing for the kinds of tools agents actually use.
The strongest theme this week is that the AI race is getting tighter at the top, while the risks at ground level keep getting more real. New models and design tools are making it easier than ever to produce professional content. At the same time, AI-driven fraud is hitting real estate hard enough that the FBI is flagging it by name. It's still a great time to lean into the productivity tools but don't get forget about the fraud that's out there.
- Jason