Trust Is Becoming Infrastructure

There has always been a lot of trust involved in a real estate transaction.

We trust that the person sending wiring instructions is actually the person we think it is. We trust that documents haven't been altered. We trust public records, signatures, photographs, property information and the identity of the people on the other side of a transaction. Most of the time we don't think about any of this as "trust." It's simply part of the machinery that allows a transaction to happen.

Artificial intelligence is going to test some of those assumptions.

I don't mean that in the usual, slightly breathless way that AI gets discussed. I'm less interested in whether someone can create a fake listing photo or write a convincing phishing email. Those things are already possible.

I'm more interested in what happens to an industry when creating something that looks authentic becomes extraordinarily cheap.

For most of our lives, there has been at least some friction involved in producing a convincing fake. Forging a document required effort. Creating a realistic photograph required skill. Impersonating someone's voice was difficult. Producing hundreds of personalized communications required people and time.

A surprising amount of our everyday trust developed in an environment where deception had a cost.

That cost is dropping quickly.

§

Creation gets cheap.
Verification gets valuable.

I've been thinking about this more recently because of work we've been doing with a title company.

Title is an interesting business because much of its value is invisible when everything goes correctly. Buyers and sellers understandably focus on the house, the financing, the inspection, the negotiations and the move. Meanwhile, an enormous amount of work is taking place to establish that the transaction is what everyone says it is.

Who owns the property?
Are there liens?
Does the person selling it have the right to sell it?
Are the documents correct?
Are the funds going where they're supposed to go?

In other words, a large part of the business is about establishing certainty.

That suddenly seems like a very modern business to be in.

As AI becomes embedded throughout real estate, I suspect we'll spend a lot of time talking about what it allows us to create. Better marketing. Faster communications. Automated documents. More sophisticated consumer experiences. Those are real opportunities.

There is another economy developing alongside all of that creation: proving that things are real.

Identity verification. Data provenance. Fraud detection. Secure communication. Document verification. Authorization. Audit trails.

These don't sound nearly as exciting as generative AI. They may turn out to be considerably more important.

§

Real estate has an unusual exposure to this problem.

A home purchase combines large sums of money, emotional consumers, complicated processes, multiple organizations and people who may never meet one another in person. Information moves between agents, lenders, title companies, attorneys, inspectors, consumers and other parties, often across different systems.

There are plenty of legitimate reasons for someone you've never met to send you an important document or ask you to take an action involving a significant amount of money.

That's a pretty attractive environment for someone trying to impersonate one of those people.

We've already spent years warning consumers about wire fraud. What's changing is the quality and scale of what can be manufactured. A fraudulent email no longer has to contain strange grammar. A fake voice doesn't necessarily sound fake. Images aren't evidence in quite the way they once were.

Eventually, "it looks legitimate" stops being a useful standard.

§

This may change what consumers value.

Technology usually promises to remove friction, and most of the time that's a good thing. Nobody wants to add unnecessary steps to a real estate transaction.

But some friction exists for a reason.

Confirming an identity takes time.
Verifying instructions takes time.
Having a person review an exception takes time.
Requiring a second form of authentication adds a step.

For years we've tended to think of those things as operational costs to be reduced. In some parts of the transaction, we may start thinking about them differently.

They are evidence that somebody checked.

There's an interesting tension here. AI will allow us to make portions of a transaction extraordinarily fast, while declining trust may require us to become more deliberate at specific moments. The best systems won't simply remove friction. They'll know where friction belongs.

§

This creates an opportunity for established real estate companies.

The technology industry has spent decades trying to remove intermediaries. Sometimes that's exactly what consumers need. But intermediaries also perform another function: they stand behind things.

A brokerage puts its name behind an agent. A title company puts its name behind a transaction. A lender verifies a borrower. An MLS creates rules around information contributed by its participants.

Those institutions aren't perfect, and a recognizable name doesn't guarantee that something is true. But the underlying function — verification by a party with something to lose — is becoming more interesting.

Reputation is a form of infrastructure.
So are processes.
So are records.
So is accountability.

Those assets aren't especially flashy, which may be why technology conversations tend to undervalue them.

§

There is a larger business question here.

If I were running a company anywhere in the real estate transaction today, I would spend some time thinking about where my organization creates trust.

Not where the marketing department says it does. Where it actually does.

What do customers rely on us to verify?
What information do we possess that other people depend upon?
Where does our involvement reduce uncertainty?
What would become riskier if we disappeared from the transaction?

There may be businesses that discover their greatest future value isn't the service they've traditionally believed they sell.

It's the certainty surrounding it.

For all the attention we're giving artificial intelligence, I suspect this will be one of its stranger effects. The easier it becomes to manufacture something convincing, the more valuable it becomes to know where something came from, who stands behind it and whether it can be trusted.

AI is going to make an astonishing amount of creation possible.

Someone is still going to have to establish what's real.