AI and Brokerage: What Is a Broker Worth When Information Stops Being Exclusive?

It is not possible to attribute the correction to a single factor. It coincided, however, with quarterly reports in which artificial intelligence dominated the dialogue with analysts. Questions from Goldman Sachs, JPMorgan, Morgan Stanley, Citi, UBS, Evercore, and William Blair focused on the technology's real scope across intermediation, productivity, and demand.

In response, companies outlined a more nuanced outlook than the stock reaction suggested. Across all four calls, one idea was repeated: artificial intelligence is being incorporated as a tool for efficiency and margin expansion, not as a substitute for the business. Operationally, this translates into lower research costs—with estimates of up to 25% in CBRE's case—reduced data acquisition and processing expenses, faster information distribution to brokers, and higher productivity per professional, allowing revenue growth without proportionally expanding headcount.

The core question, however, is not whether artificial intelligence improves brokerage efficiency but whether it can replace it.

The companies argue it cannot. They maintain that their activity does not consist of intermediating standardized assets, but of structuring large and complex transactions, in which strategic judgment, negotiation, and fiduciary responsibility remain decisive. In that arena, information is not just processed data; it is interpretation, experience, and relationships. That is where the sector believes the barrier to automation lies.

CBRE summarized this by noting that it is not "selling $2 million condominiums," but executing highly complex transactions. Similarly, JLL said it does not see external competitive pressure; Cushman acknowledged that artificial intelligence will produce winners and losers but dismissed the prospect of broad disintermediation; and Newmark admitted greater vulnerability in segments perceived as commoditized, limiting that risk to mid- or lower-market transactions.

In general terms, they acknowledged limited vulnerabilities: partial automation in valuations and research, greater exposure in repetitive tasks, and possible pressure in certain Class B office assets. There have been no warnings about structural fee compression, broker reductions, guidance adjustments due to technological risk, or loss of mandates to native artificial intelligence platforms.

Digitalization itself, by contrast, is presented as generating new infrastructure needs. The expansion of data centers and computing capacity associated with artificial intelligence implies additional demand for space, energy, and specialized personnel. In that context, rather than disappearing, competition shifts: strategic assets change, but the activity itself does not necessarily.

That redistribution, however, does not resolve the core debate. Even without replacing the intermediary, technology can narrow competitive distance by making information processing and circulation more accessible. And when access ceases to be differentiating, value shifts toward the ability to interpret, structure, and negotiate. Thus, the risk is not the broker's disappearance, but the reconfiguration of the asset that historically captured value.

Ultimately, what this episode reveals is not a consummated disruption, but a change in how the market values the sector. Artificial intelligence has moved from being an operational matter to becoming a structural variable in the assignment of multiples, in a context where, if differentiation is perceived as less solid, valuation adjusts before financial results do.

For commercial real estate, this means that the discussion around technology has moved from the margins to the center of risk analysis. The question is no longer whether artificial intelligence improves processes, but how it alters the stability of the rents the sector once considered structural.

For now, the February adjustment was not a conclusion. It was a turning point.

More information and commercial real estate market data are available at SiiLA or at contacto@siila.com.mx.

***

¹ The positioning of the consensus did not suggest structural deterioration. At CBRE, 79% of recommendations were "Buy," with a median price target of $187—about 27% above the prevailing level—and a P/E of 38x. At JLL, the consensus implied 24.5% upside (P/E 18.97x). Cushman & Wakefield showed a median target 39% above the market price, while at Newmark, 88% of recommendations were "Buy" and the median target suggested upside of roughly 45%.