AI Real Estate Agent Closes Its First Deal

On October 15, seller Ms. Wang and buyer Mr. Liu officially signed a sales contract for a second-hand apartment in Phase 6 of Zhuhai Huafa New City and completed the ownership transfer. The home, sold for a total of 1.95 million yuan, is the first transaction in China led by an AI real estate agent. AI agent "Yijuxiaoxin" handled the core stages of the deal online, including listing, client matching, property verification, providing a community reference transaction price, and online price negotiation. Human agents appeared only in offline on-site viewings, professional contract notarization, and transfer services. The deal had one notable feature: the AI agent charged no commission. The seller passed on the full 40,000 yuan commission that would traditionally have been paid to an agency to the home's next owner. Moving from a "supporting" role to the core of transaction matching, AI has taken a key step forward. Currently, AI applications in China's real estate brokerage sector follow two fundamentally different paths. The mainstream model uses AI as a tool to improve efficiency and service, assisting agents in their work. AI acts as an add-on to agencies, leaving the existing business model, commission rules, and transaction decision-making in agents' hands. Beike, 58.com's Anjuke, and small and medium-sized agency platforms all use it extensively. Representative products include Beike's AI workbench, Anjuke's intelligent chat, and various third-party tools for property copywriting and digital humans. Tools already exist for property verification, information structuring, valuation, listing content creation, managed customer service, lead scoring, conversion prediction, and tax calculation. They are mature and widely used, but AI does not independently connect buyers and sellers and cannot complete matchmaking without agents. The other path is the independent AI intermediary agent. Here, AI is no longer an agent's tool but a neutral intermediary in its own right, with independent matchmaking authority. It can initiate conversations between buyers and sellers directly, without agents relaying messages. The goal is to complete supply-demand matching, two-sided matchmaking, and price negotiation directly, connecting buyers and sellers. However, this model is currently rarely used and remains in an exploratory stage, with the transaction process and business path not yet fully closed-loop. The two models differ in one major respect: in the first, AI mainly serves real estate agents and other intermediaries, while the second serves individual owners and home buyers (C-end users) directly. Unlike new homes, second-hand home buyers and sellers are typically dispersed individuals, so transactions depend more heavily on intermediary services. As a result, the transparency and commission rates of second-hand brokerage services have long drawn attention. Agents pushing prices down or up, or concealing transaction information, often