A Miami father sold his house in five days using ChatGPT. He cleared $954,800, roughly $100,000 above what local agents had quoted him. He skipped the broker, kept the commission, and told the press he’d do it again.

It’s a good story. It’s also a dangerous one to overgeneralize.

Robert Levine’s experiment was real, and the results were genuinely impressive. He used an AI chatbot to set pricing strategy, time the listing, write marketing copy, and field offers. Within 72 hours of going live, five buyers had raised their hands. By the end of the weekend, he had a signed contract. The tool even flagged which rooms to repaint for maximum return, and it was right. For a four-bedroom house in a liquid, data-rich American market where comps are plentiful and buyers are predictable, the approach worked.

But here is what Levine actually proved: AI is very good at processing information that already exists in a structured, accessible format. The MLS has that. Miami’s residential market has that. Thousands of comparable sales, standardized disclosures, public records, and a relatively uniform buyer pool all feed a machine that can, in fact, tell you to list on a Wednesday.

Now apply that same logic to a 16th-century palazzo in the Oltrarno district of Florence, or a cliffside villa above the Amalfi Coast, or a converted masseria in Puglia. See how far the algorithm takes you.

The Data Desert Problem

Italy’s luxury property market runs on relationships, not databases. There is no national MLS. Listings are fragmented across private networks, notarial records, and word-of-mouth channels that have operated the same way for decades. Asking prices are often theater; actual transaction values are closely guarded. The buyer who paid €3.2 million for a Chianti estate last spring didn’t announce it on Zillow.

This opacity is not a flaw. For serious buyers, it is actually the point. The scarcity of clean data keeps certain markets insulated from the kind of algorithmic pricing pressure that compressed Miami margins during the pandemic boom. When information is restricted, the people who hold it have real power. That power belongs to agents, notaries, local developers, and the small circle of intermediaries who move between markets and speak both languages, legal and cultural.

ChatGPT cannot cold-call a Florentine seller’s family lawyer. It cannot sit across a table in Milan and read whether a developer is genuinely motivated or simply testing the water. It does not know which Venetian neighborhood association will delay a renovation permit for three years, or which won’t.

Why Americans Are Moving Anyway

None of this has slowed American appetite for Italian property. Quite the opposite. Since 2021, dollar-denominated buyers have emerged as one of the most active forces in Italy’s upper market, drawn initially by currency advantage and kept by something harder to price: the quality of life that the country quietly offers people with the resources to access it.

The numbers support the momentum. Prime residential values in Florence and the surrounding hills have climbed consistently, with international demand absorbing inventory that would have sat in prior cycles. Rome’s historic center has seen renewed institutional interest. The Sicilian coast, once considered too remote for serious luxury investment, is now attracting buyers who spent the pandemic rethinking what proximity actually means.

What’s driving this is not just aesthetics. American buyers who have gone through the process describe a distinct shift in how they think about asset allocation. A Tuscan farmhouse is not a vacation home. It is a euro-denominated hard asset with rental income potential, a hedge against dollar volatility, and a generational transfer vehicle that benefits from Italy’s relatively favorable inheritance framework. The lifestyle component is real, but sophisticated buyers are no longer using it as the primary justification.

The Tool and the Expert Are Not the Same Thing

Back to Levine, and to the honest lesson his story contains.

He used AI to move faster and smarter through a process he largely understood. He knew his market. He knew his product. He had access to all the relevant data. The chatbot organized and accelerated decisions he was already capable of making. He also, it is worth noting, brought in a lawyer for the final paperwork. He did not fully replace human judgment. He augmented his own.

That is exactly right. And it is also the ceiling of what AI can do in real estate, particularly at the luxury end of the market and particularly across international borders.

The cross-Atlantic buyer, whether an American looking at Lombardy or an Italian family exploring South Florida, is not navigating a data-rich, standardized system. They are navigating two legal frameworks, two tax regimes, two currencies, and two entirely different cultures of negotiation. The stakes are higher. The transactions are less liquid. The margin for error is smaller.

In that environment, the most valuable thing is not an algorithm. It is a person who has done this before, who knows who to call on both sides, and who has enough standing in both markets to get a straight answer.

The Bridge Still Has a Human at Each End

Columbus International Real Estate was built precisely for this corridor. Headquartered at Rockefeller Center in New York with offices in Miami, Milan, and Florence, the firm operates less like a traditional brokerage and more like a private intelligence network for buyers and sellers who are serious about moving between markets. The firm’s focus on the Italy-U.S. corridor is deliberate: it reflects where the most complex, highest-value cross-border transactions are actually happening. For Italian investors navigating American markets or American buyers pursuing Italian acquisitions, Columbus International functions as the expert guide who already knows the terrain. Reach them at info@columbusintl.com.

The algorithm will keep improving. It will price Miami houses faster, write better listing descriptions, and eventually absorb more of the transactional friction that currently costs sellers three percent. None of that is bad news.

But the conversation that closes a deal on a Florentine courtyard apartment still starts with a phone call to someone who knows the owner personally.