Several large escrow operations were supposedly being shut down by their corporate owner. Hundreds of employees potentially stood to lose their jobs. The companies would stop accepting new escrows on a specific date, existing transactions would be allowed to close, and employees would remain temporarily to help customers through the transition. Only one problem. It wasn't true. 

And it wasn't a vague rumor either.

 

The information came in the form of what purported to be a written copy of an announcement to employees. It identified the affected operations and their affiliated brands. It discussed severance, health and welfare benefits and opportunities for employees to apply for other positions within the parent company. It even directed employees to an internal Human Resources SharePoint site for additional information.

 

There were specific dates. There were specific procedures. There were specific corporate relationships.

 

And almost everything surrounding the story checked out.

 

Except, apparently, the story itself.

 

 

The information reached me after an executive I trust in a large real estate brokerage had it sent to him, and he distributed it within the company. 

 

That alone gave it credibility. It was actually a "Wow! moment." The names of the companies involved are well known within the industry in Southern California. In fact, they are some of the largest escrow firms in the area. 

 

The story wasn't something I found posted anonymously on social media by someone with 14 followers and using an egg for a profile picture. It had reached senior people within the real estate industry and was detailed enough to be taken seriously.

 

So I started looking into it.

 

The affected companies really were owned by the corporation identified in the announcement.

 

The corporate relationships between the various brands were correct.

 

A little investigation allowed me to determine that the parent company had been investing heavily in AI technology designed to automate escrow and settlement work. So the closures and the layoffs that would come with them appeared to make economic sense for the company.

 

It really had developed an artificial-intelligence-driven settlement platform.

 

It really had tested that technology in actual real estate transactions.

 

It really had publicly discussed dramatically increasing the percentage of escrow work that could eventually be automated.

 

And it really had announced plans to deploy that technology much more broadly over the next couple of years.

 

They were rolling out a more efficient, less costly, more profitable escrow system... or so I thought. The alleged shutdown made a lot of sense.

 

 

The potential implications were enormous.

 

These were white-collar jobs, many of them held by experienced people earning good incomes.

 

If hundreds of traditional escrow positions were disappearing while the corporate parent was simultaneously rolling out technology capable of automating much of the work those employees performed, this wasn't simply a story about a slow real estate market. It potentially offered an early glimpse of what artificial intelligence could do to an entire industry.

 

The economics weren't difficult to understand. Escrow remains surprisingly labor intensive. Officers, assistants and processors perform enormous numbers of repetitive tasks associated with moving a real estate transaction from an accepted offer through closing. Automate enough of those tasks and one employee can handle substantially more transactions.

 

That means fewer employees.

 

It also means lower costs and potentially higher profits.

 

There was another interesting question. Would those savings eventually be passed along to buyers and sellers through lower escrow fees? There was little indication that they would, at least initially.

 

That led to a larger story about artificial intelligence and employment. Escrow wouldn't be unique. Insurance, banking, accounting, mortgage processing, legal services and dozens of other white-collar industries contain similar repetitive workflows.

 

This appeared to be a remarkably good case study of something we're likely to see repeatedly as AI adoption accelerates.

 

There was just one problem.

 

While I had already finished writing the story, I couldn't independently confirm that the companies were actually closing. And I wasn't going to publish without that verification.

 

 

I kept looking for a public announcement from the corporate parent.

 

Nothing.

 

I looked for reporting in real estate and title-industry publications.

 

Nothing.

 

Then I started looking for California WARN notices.

 

California's WARN Act generally requires qualifying employers to provide advance notice of certain mass layoffs and facility closures. If hundreds of people really were about to lose their jobs, it seemed reasonable to expect some sort of paper trail.

 

I couldn't find one.

 

That didn't necessarily disprove anything. There are exceptions, filing delays and questions about how employees at different locations are counted. The notices could also have been filed under corporate names that weren't immediately obvious.

 

So I waited.

 

Meanwhile, the websites of the supposedly doomed companies remained online.

 

More importantly, they continued looking like businesses that wanted customers.

 

One was still telling visitors to open an escrow. In fact, it stil is today even though the announced data for no new escrows has already passed. I checked.

 

Another continues inviting people to get started. Again, I checked.

 

Even more strangely, employment advertisements remain online. One of the companies is inviting people to send in résumés to join its growing team.

 

That doesn't  make much sense for businesses that supposedly have stopped accepting new customers and are preparing to eliminate their entire workforces.

 

But websites can be neglected. Job postings can remain online after positions have been canceled. Corporate communications departments aren't always particularly quick about updating things.

 

None of that was enough for me to declare the original information false.

 

 

There is a certain irony here.

 

At ACCESS, we're a nonprofit organization that, among other things, spends a considerable amount of time educating people about scams and fraud.

 

One of the most basic lessons in avoiding scams is also one of the simplest: Verify the information independently.

 

That principle shouldn't apply only when somebody receives a suspicious text message asking for a credit-card number.

 

It should apply to information we publish as well.

 

And that has become increasingly important because modern publishing rewards speed.

 

A story appears. Someone posts it. Someone else repeats it. A third person writes an article based upon the first two posts.

 

Within hours, dozens of websites can be reporting essentially the same information.

 

It looks independently confirmed because there are suddenly 20 sources.

 

Except sometimes all 20 sources trace back to one source that nobody bothered to verify.

 

I had a good story.

 

I had a detailed document.

 

I had independently verified many of the facts surrounding it.

 

What I didn't have was confirmation of the most important fact: Were these companies actually shutting down?

 

 

Finally, I decided to try something considerably less sophisticated than artificial intelligence, corporate filings or internet research.

 

I picked up the telephone.

 

I called one of the escrow companies supposedly being shut down.

 

I asked whether they were still accepting new escrows.

 

Yes.

 

Then I asked directly whether the company was going out of business.

 

No.

 

That was pretty much the end of my already written article.

 

And I was very glad I hadn't published it.

 

 

There is another side to this story that bothers me considerably more.

 

Someone apparently went to substantial effort to create or circulate information that was convincing enough to fool experienced people within the real estate industry.

 

Consider what could happen to an escrow company if Realtors and brokers throughout Southern California were told that it was about to close.

 

Would you open a new escrow there?

 

Probably not.

 

Would you advise your client to send a substantial earnest-money deposit to a company you believed would cease operations within weeks?

 

Probably not.

 

Would lenders, title representatives and other professionals begin steering transactions elsewhere?

 

Quite possibly.

 

Escrow is particularly vulnerable to this kind of rumor because transactions last for weeks and involve significant amounts of money. Customers don't simply care whether a company is operating today. They need confidence that it will still be operating when their transaction closes.

 

A sufficiently credible false report that an escrow company was shutting down could therefore become self-fulfilling.

 

New orders could disappear.

 

Employees could start looking for other jobs.

 

Experienced escrow officers might leave.

 

Competitors could begin recruiting those employees and pursuing their customers.

 

Business relationships built over decades could be damaged in days.

 

Even after the rumor was disproved, some of the damage might remain.

 

That raises an obvious question.

 

Why would someone do this?

 

I don't know.

 

And I'm not going to speculate about who created the information or what their motive might have been without evidence.

 

That would rather defeat the point of this article.

 

 

We tend to imagine misinformation as something relatively easy to spot. A badly written email. An obviously manipulated photograph. A ridiculous social-media post.

 

The more dangerous version looks nothing like that.

 

It contains real names. 

 

Real companies.

 

Real corporate relationships.

 

Real technology.

 

Real business trends.

 

Realistic dates.

 

Professional language.

 

And just enough truth to make the false part seem obvious.

 

That is what made this particular information so convincing.

 

Once I confirmed that the corporate parent really was investing heavily in automated escrow technology, the supposed shutdown stopped looking surprising. In fact, it began looking almost inevitable.

 

That was precisely the danger.

 

Every additional fact I verified made me more confident in a conclusion I hadn't actually verified. And in the age of artificial intelligence, creating something this convincing is only going to become easier.

 

 

There is a lesson here for anyone who publishes information, whether it's a major news organization, a blogger, a corporate executive or someone with a social-media account.

 

Speed isn't verification.

 

Detail isn't verification.

 

Twenty websites repeating something isn't necessarily verification.

 

Even confirming most of the facts surrounding a story doesn't confirm the story itself.

 

Sometimes you have to slow down long enough to ask the simplest question: How do I know this is actually true?

 

In this case, after weeks of research into corporate structures, employment numbers, regulatory filings, artificial intelligence, escrow automation and business strategy, the most useful investigative tool turned out to be one that Alexander Graham Bell would have recognized.

 

I called the company.

 

They answered.

 

And they were still in business.

 

 

Comments

Search

3.25 Copyright (C) 2007 Alain Georgette / Copyright (C) 2006 Frantisek Hliva. All rights reserved."