Google’s algorithms know me well and appreciate my fascination with consumer scams. My news feed is routinely chock full of stories about people losing their life savings to fraud, a subject I’ve written about before, noting that customers of JPMorgan Chase seemed particularly susceptible. One might expect a bank run by Jamie Dimon, routinely portrayed as America’s most brilliant banker, to set the gold standard for protecting customers.
Yet The Wall Street Journal recently reported that federal prosecutors reviewed claims that Chase executives ignored serious deficiencies in the bank’s antifraud program and improperly denied more than $100 million in reimbursements to customers whose money had been stolen. The review stemmed from a whistleblower report filed by the bank’s former head of scam prevention.
Chase disputes the allegations, insisting it acted appropriately and within the law. Perhaps. But complying with the bare legal minimum isn’t necessarily the same as doing everything reasonably possible to protect customers.
It’s alarming that America’s largest bank, with virtually unlimited resources devoted to fraud prevention, is facing allegations that it failed to protect customers from increasingly sophisticated scams. Consumers need a different kind of protection, one that’s available before they become victims. Artificial intelligence is already sufficiently developed to fill that role.
Kimberly Fudge’s Jury Duty Misadventure
The Journal recently profiled Kimberly Fudge, a 38-year-old Florida finance manager who studied mechanical engineering and mathematics. Scammers posing as sheriff’s deputies convinced her she had missed federal jury duty and faced arrest. Over nearly ten hours they kept her on the phone, threatened her with jail, claimed they were monitoring her location, sent fake legal documents and directed her to deposit cash into cryptocurrency kiosks as “bail.” They knew her Social Security number, details about her family, and appeared able to monitor activity on her phone.
The scam finally unraveled after Fudge’s husband became suspicious, asked ChatGPT about his wife’s bizarre request for money, and then called 911. By then the couple had lost $25,000.
Reading the story, I found myself wondering how anyone with a modicum of intelligence could possibly fall for such an obvious scam.
Then I caught myself.
A week earlier, I had nearly become one of those stories.
“There’s a Lien on your Property”
Late one morning last week, while already irritated that I’d let too many administrative chores pile up, I received an unsolicited call from someone claiming to be a mortgage broker.
“I was reviewing your property and noticed there’s a lien against it,” he said. “Were you aware of that?”
I wasn’t.
“Who placed the lien?” I asked.
“I’ll need to do more research,” he replied. “Would you like me to investigate?”
The conversation immediately struck me as odd.
The supposed broker knew my address and cellphone number yet couldn’t tell me who had supposedly placed the lien. He also lacked the polish and professionalism. I was suspicious enough to tell him, in as ominous a tone as I could muster, that I’d first be consulting my lawyer.
Still, the call unsettled me.
A year earlier I’d had a dispute with a contractor who threatened to place a lien on my property unless I paid him immediately. I caved. Had he filed one anyway?
I asked ChatGPT.
The software immediately questioned the broker’s story. If he knew there was a lien, why couldn’t he identify who recorded it? ChatGPT also advised me that if I had a mortgage, my lender would already have a lien against the property, which is entirely routine. It told me not to be alarmed that the caller knew my address and cellphone number because both are readily available through commercial databases.
As a precaution, it suggested I verify any lien through official public records rather than relying on an unsolicited caller. It also recommended locking my credit files, advice I’d previously followed.
I Googled California property records.
The first result was a website called California Public Records Data. Like most people, I assumed Google’s top result was legitimate. I entered my name and address, and a few minutes later the site proudly informed me my report was ready.
The price?
One dollar.
That sounded reasonable enough. I was prepared to risk a buck.
But I hesitated before entering my credit card.
So I asked ChatGPT again.
The software pointed out something I’d completely missed in my haste to find out whether someone had placed a lien on my property. I wasn’t buying a one-dollar property report. I was agreeing to a two-year subscription costing $30 a month, for a total commitment of roughly $720.
The website said I could cancel at any time. The Better Business Bureau website and other online forums, however, contained complaints from customers alleging cancellation wasn’t nearly so simple.
My cousin Rob works in real estate, so I emailed him asking whether he had access to property records.
Within minutes he sent me a link confirming there was indeed a lien against my home.
It belonged to Chase.
My mortgage.
Consumer Protection, Reinvented
Artificial intelligence has received an avalanche of deserved criticism, particularly over its potential to eliminate jobs, a prospect that leaves many CEOs positively drooling. But remarkably little attention has been paid to AI’s potential as a consumer protection tool, filling a market void that appears to be growing wider every year.
Americans reported losing a staggering $3.5 billion to imposter scams in 2025, according to the Federal Trade Commission, nearly triple the amount reported in 2020. Imposter scams accounted for nearly one in three fraud reports. They arrive by text, email, phone calls, social media, search engine results and increasingly sophisticated fake websites.
Banks, credit card companies, regulators and law enforcement are all supposed to play important roles in protecting consumers. The problem is that almost all of them step in after the damage has been done. AI offers consumers something we’ve never really had before: an intelligent consumer advocate sitting beside us before we click “Accept,” enter our credit card number or answer a threatening phone call.
For decades, corporations have hired armies of lawyers, engineers, marketers and behavioral scientists to gain advantages over consumers. AI may become the first technology that gives ordinary people access to comparable expertise.
Consider Texas Attorney General Ken Paxton’s privacy lawsuit against General Motors.
According to the complaint, GM electronically presented customers with more than fifty pages of disclosures, product descriptions and privacy notices before obtaining their consent to collect and sell data about their driving habits.
How many customers actually read those fifty pages? More importantly, how many could realistically understand them?
Imagine uploading those documents into an AI assistant and asking one simple question:
“What am I really agreeing to?”
Within seconds the software could identify provisions involving data collection, arbitration clauses, recurring charges, privacy waivers, cancellation rights and other terms consumers routinely overlook because few people have the time or legal training to decipher fifty pages of corporate disclosures.
That’s not practicing law. It’s translating legalese into English. For millions of routine consumer decisions, that’s more useful than hiring a lawyer.
Imagine having an AI consumer advocate that already understands your finances, subscriptions, insurance policies, travel preferences, health records and tolerance for risk.
It answers unknown callers before your phone even rings.
It screens text messages and emails for fraud.
It reviews contracts before you sign them.
It explains terms and conditions in plain English.
It tells you whether a “free” offer is actually free.
It reminds you that a “one dollar” property search really costs $721.
And perhaps most importantly, it never gets intimidated, distracted or rushed.
My favorite feature would be authorizing it to talk directly to corporate chatbots. Instead of spending an hour explaining my problem to an airline, insurance company or bank, I’d simply tell my AI what happened and let the two chatbots sort it out. Their chatbot would finally meet mine.
Of course, my chatbot would also be experiencing “longer than normal wait times due to unusually heavy call volumes.”
If the matter couldn’t be resolved, my AI could organize supporting documents, draft a demand letter, prepare a small claims filing, complete regulatory complaints and tell me whether I had a legitimate grievance or was simply wasting my time.
That’s not science fiction. Much of the underlying technology already exists. ChatGPT, Gemini and Claude can already explain contracts, identify suspicious requests and summarize legal disclosures.
What’s missing is integration. Consumers still must remember to stop, copy a document into a chatbot and ask the right question. A true AI consumer advocate would do that automatically, the way antivirus software quietly scans your computer without waiting to be asked.
Honest companies should welcome that future. Businesses that compete through transparency and fair dealing have little to fear from informed customers. The ones that rely on teaser pricing, unread contracts, confusing disclosures and customer exhaustion won’t.
Those are the companies consumers should stop rewarding with their business.