Fighting AI with AI

Summary

Podcast Summary: "Fighting AI with AI"

The Indicator from Planet Money – October 6, 2025
Host: Darian Woods (with reporting from various NPR staff)
Theme:
This episode delves into the rapidly evolving landscape of AI-driven voice fraud — specifically, scams that use cloned voices to commit crimes — and examines how companies and banks are fighting artificial intelligence… with artificial intelligence.

---

Main Theme & Purpose

The episode explores how criminals increasingly use AI to replicate voices and scam both customers and institutions, focusing on the vulnerabilities in banking and the countermeasures being developed to detect and block these threats. It features firsthand anecdotes, expert commentary, and a look at the technological arms race between scammers and defenders.

---

Key Discussion Points & Insights

1. How Voice Deepfakes Are Used for Scams

• Darian Woods pranks his colleague Angel Carreras with an AI voice, simulating a common scam technique of impersonating a coworker to request gift cards.
• Angel quickly detects the ruse, exposing the growing sophistication but also the limitations of AI-generated voices (00:27–00:57).
• Commentary highlights how millions of Americans have already lost substantial amounts to AI voice scams, with convincing audio easily tricking victims—especially when urgency is invoked (01:17–01:40).

2. Banks on the Front Line

• Darian Woods explains banks are major targets of AI voice fraud due to the “literal” monetary stakes (03:05–03:10).
• Mark Guapozeski, Chief Information Officer at PNC Bank, discusses the constant pressure to strengthen security as criminals evolve with digital banking (03:22–03:37).
• A typical scam involves recording a victim’s voice for a few seconds to create an AI clone, then using it to bypass phone-based voice ID systems (03:37–03:49).
• Quote (Mark Guapozeski, 03:49):
  “Because it’s so easy now to reproduce your voice, you really can’t rely on any one vector to say, okay, I’m just going to accept it’s you because I hear you.”

3. Rise of ‘Reality Defender’ and Detection Tech

• Ben Coleman, former Goldman Sachs employee and co-founder of Reality Defender, has been tackling AI fraud risk since 2021, anticipating the coming “tsunami of fraud” despite lacking terms like “deepfake” at the time (03:59–04:19).
• Coleman explains their detection relies on “inference,” identifying subtle features in audio that signal AI generation, undetectable by humans (05:00–05:13).
• Quote (Ben Coleman, 05:13): “There’s indicators of AI, which means that the information is, yes, it’s yours, but it’s being used by somebody who’s not you.”
• Reality Defender is now used by the majority of the top 20 banks.

4. Limitations and Need for Multi-Factor Security

• Ben Coleman warns that reliance on voice biometrics—“your voice is your password”—is dangerously outdated, advocating for stronger multilayered systems (05:42–05:56).
• Mark Guapozeski details PNC’s approach: using location, caller device, security codes, and other data points to supplement voice authentication (06:11–06:49).
• Quote (Mark Guapozeski, 06:11): “If you’re only using the one dimension, there’s risk in everything… you always want to have layers of security.”

5. Target Shift: From Banks to Customers

• Reality Defender and similar tools have hardened bank systems, shifting fraudsters’ focus to exploiting bank customers directly (07:11).
• Scammers call customers pretending to be the bank, urging them to move funds or invest in crypto/gold, but Mark warns:
  Quote (Mark Guapozeski, 07:36): "We will never ask you to move your money. If we think an account has been compromised, the bank will move your money for you, you know, within the bank."
• He tells how fraudsters build a sense of urgency and notes the importance of verifying the institution’s identity.

6. Practical Defenses: Family Safe Words

• Mark shares that his family uses a “safe word” in case anyone calls asking for money due to supposed emergencies—a simple but effective tactic (08:06).
• Quote (Mark Guapozeski, 08:06): “We all know if there’s ever a situation where somebody is either really in trouble and asking for money, ...we will ask for the safe word.”

7. The Need for Broad AI Detection Regulation

• Ben Coleman asserts that banks alone won’t be enough; all online content should carry an indicator if it has been AI-generated, from voice to video (08:27–08:44).
• He explains their push for regulations, having demonstrated deepfakes in Congress, even mimicking Senators’ voices to prove the threat (09:13–09:24).
• Quote (Ben Coleman, 08:53): “I think we’re gonna look back and say, I can’t believe there was a time when we didn’t have automated deepfake detection. Our challenge is just the technology is moving quicker than regulations.”

8. Real-World Scams and the Societal Stakes

• The hosts note celebrity impersonation scams (such as fake pro golfers on Instagram) are proliferating (08:44–08:53).
• Darian notes even The Indicator has been impersonated by scammers, warning listeners to check for official NPR email addresses.

---

Notable Quotes & Memorable Moments

• Darian’s AI prank on Angel:
  Angel Carreras (00:27): “You sound like AI.”
  Angel Carreras (00:55): “Too sloppy.”
• Mark Guapozeski on risk:
  (03:49): “Because it’s so easy now to reproduce your voice, you really can’t rely on any one vector…”
• Ben Coleman’s vision for detection:
  (08:53): “I think we’re gonna look back and say, I can’t believe there was a time when we didn’t have automated deepfake detection.”
• Use of safe words:
  Mark Guapozeski (08:06): “We all know if there’s ever a situation where somebody is…asking for money, ...we will ask for the safe word.”

---

Important Timestamps

• 00:11–01:05: Darian pranks Angel with AI-generated voice
• 01:17–01:40: Overview of AI scam prevalence and losses
• 03:05–03:10: Banks’ vulnerability explained
• 03:22–03:49: Mark Guapozeski (PNC Bank) on fraud evolution
• 04:19–05:13: Ben Coleman introduces Reality Defender’s approach
• 05:56–06:49: Risks with single-vector authentication, move to multi-factor
• 07:11–07:59: Customer-targeted scams, advice from PNC
• 08:06–08:24: Safe word practical tip
• 08:44–09:24: Regulatory need; demonstration in Congress
• 09:39: Reminder for listeners about official Indicator contacts

---

Tone & Style

The episode blends wit, urgency, and expert commentary, keeping jargon minimal and translating complex technical threats into relatable, everyday risks. The hosts simultaneously entertain (via pranks and banter) and educate, making the topic both clear and memorable.

---

Conclusion

With AI-generated voices now capable of fooling both ears and security systems, defenses must be both technical (like Reality Defender’s detection software) and practical (such as safe words and awareness). The episode emphasizes multilayered security, continued vigilance, and the need for regulation as technology leaps ahead of policy.

---

Next episode teaser: A look at what’s “supercharging” data breaches.

Transcript

[Mark Guapozeski]

Npr.

[Darian Woods]

So, Waylon, the other day I decided to give our colleague Angel Carreras a little test.

[Angel Carreras]

Hello, Darian.

[Ben Coleman]

Hey.

[Angel Carreras]

Hey. Sorry, do you have headphones on?

[Darian Woods]

Angel, I need your help with something real quick.

[Angel Carreras]

What's up?

[Darian Woods]

I need your help with something real quick.

[Angel Carreras]

You sound like AI.

[Darian Woods]

Would you be able to go out and buy some gift cards for me?

[Angel Carreras]

You sound like, I don't have time today.

[Darian Woods]

It's a surprise for our colleague.

[Angel Carreras]

Darian, I'm so sorry. You're not fooling me. This is AI.

[Darian Woods]

If you could just go out and buy, like, $200 in gift cards, that'd be awesome.

[Angel Carreras]

Okay. Actually, you've convinced me. Where should I get these? Dare.

[Ben Coleman]

Dai.

[Angel Carreras]

Dai. Darian, Come on, quick. On your feet. Darian.

[Ben Coleman]

Where?

[Darian Woods]

Angel, I have to admit something to you.

[Angel Carreras]

Too sloppy.

[Darian Woods]

This is not my real voice.

[Angel Carreras]

Too sloppy.

[Darian Woods]

I am a deep fake. Gosh, Angel. I got you two.

[Angel Carreras]

You got. You got nothing. You got nothing.

[Narrator/Reporter]

Oh, my gosh. Angel was so clever. If you had called me, I would have fallen for it immediately. Yeah, I would have been like, where do you want the gift cards from? Gap. Do you need a Gap gift card?

[Darian Woods]

No, I know. And so, yeah, depending on the person, depending on the situation, like, maybe it'd been urgent, like a cousin calling from a hospital. This could have really fooled somebody.

[Narrator/Reporter]

Yeah. A lot of people are falling for these kinds of audio deep fakes. Like millions of Americans have lost money to a scam call that uses an AI voice. And the losses from these scams can be in the thousands of dollars.

[Darian Woods]

But it's not all hopeless. Businesses are using AI to fight AI. This is the Indicator from Planet Money. I'm Darren Woods. This week on the Indicator, we're going to bring you a special series on the evolving business of crime.

[Narrator/Reporter]

Today's show, defending against AI Voice clones. We speak with the company working to ferret out AI deepfakes. And we learn how banks are fortifying themselves in an age where anyone can sound like anyone else.

[Indeed Sponsor]

Support for NPR and the following message come from. Indeed. Just realizing your business needed to hire someone yesterday, speed up your hiring right now with Indeed. Claim your $75 sponsored job credit now at indeed.com indicator terms and conditions apply.

[Edward Jones Sponsor]

Support for NPR. And the following message come from Edward Jones. A rich life isn't always a straight line. Unexpected turns can bring new possibilities. With a hundred years of experience navigating ups and downs, Edward Jones can help guide you. Let's find your rich together. Edward Jones. Member, sipc this message comes from Schwab. Everyone has moments when they could have done better. Same goes for where you invest. Level up and invest smarter with Schwab. Get market insights, education and human help when you need it.

[Darian Woods]

Banks are a big target of AI voice fraud.

[Narrator/Reporter]

I mean, that's where the money is, literally.

[Darian Woods]

And so if you're a chief information officer at a bank, you're constantly checking your websites and phone line systems for any vulnerabilities. That's what Mark Kwapozesk is doing at PNC bank.

[Mark Guapozeski]

Fraudsters constantly bang on every door trying to find any crack in essentially our armor or our moat, you know, around the bank. As banking has gone more digital, obviously the criminals have followed there.

[Narrator/Reporter]

One scheme calls people up, it records you talking for several seconds and then it turns that into a cloned AI voice and uses that to bypass bank's voice verification on the phone.

[Mark Guapozeski]

Because it's so easy now to reproduce your voice, you really can't rely on any one vector to say, okay, I'm just going to accept it's you because I hear you.

[Darian Woods]

One person has been on this problem for years. Ben Coleman used to work for Goldman Sachs, and there he saw the early stages of these kinds of frauds. So in 2021, he co founded a tech company that would try to protect big institutions like banks from voice fraud. He just didn't have the language for it yet.

[Ben Coleman]

But we didn't have the buzzword of Deepfix, we didn't have generative AI, we didn't have ChatGPT or anything. So we said we can detect AI avatars and virtual humans, which are about to be this huge kind of tsunami of fraud.

[Darian Woods]

How did you see this coming?

[Ben Coleman]

I just assumed if I was a hacker, what would I be doing? How do I do more hacking?

[Narrator/Reporter]

That's what I ask myself every day.

[Darian Woods]

Jerry Blackout hacker Waylon Wong Watch out.

[Narrator/Reporter]

Ben thought that voice verification was this huge vulnerability for banks. If the bank verified your voice on the phone, you could do a wire transfer or reset the password and gain complete control. He named his company Reality Defender.

[Ben Coleman]

We're doing what's called inference, which is looking for different features that probabilistically indicate that AI was used.

[Darian Woods]

An AI voice has a particular harmonic structure that the human ear doesn't hear, but Reality Defender software can detect.

[Ben Coleman]

There's indicators of AI, which means that the information is, yes, it's yours, but it's being used by somebody who's not you.

[Darian Woods]

So your company is almost like one of these AI detection websites where teachers might put a student's essay and say, is this AI, but you'll turnitin.com?

[Ben Coleman]

Yeah.

[Narrator/Reporter]

Ben says the majority of top 20 banks use reality Defender software and many use other services like Ben's. But he says that banks should step up even further.

[Ben Coleman]

Unfortunately, many institutions, not only banks, but also government organizations, insurance companies, media organizations, are still using what are called voice biometrics, which is a short way of saying your voice is your password.

[Darian Woods]

Ben thinks banks should just remove seeing a voice as a kind of password entirely. We asked Mark Guapozeski at PNC bank why it was still using voice id. Are there risks to PNC using voice authentication technology?

[Mark Guapozeski]

I think if you're only using the one dimension, there's risk in everything. There'd be risk in accepting a driver's license, for example, somebody walking into your branch, which is why you're starting to see a lot more technologies that are looking for multi factor authentication or even just those other signals. So I would sort of say you always want to have layers of security. And that's probably the key thing, you know, that we're always looking at is which how well do we feel we have our various layers covered. And then you're able to learn where you might be overusing one of those signals and adjust.

[Narrator/Reporter]

So it's not just your voice the banks are looking at, but also your location, the device you're calling from, details like your birthday, a text message, verification code, all kinds of things. In fact, Mark says criminals flock to where the weakest defenses are. And thanks to software like Reality Defender, the greatest vulnerability isn't with bank phone lines, it's with the customers.

[Mark Guapozeski]

So it's a little bit in reverse. How do you know that the company that just called you is actually pnc? We've spent a lot of time and money with, you know, the telecom carriers, different technology companies, so that if somebody's trying to spoof one of PNC's numbers, it gets blocked and it never gets delivered to you.

[Darian Woods]

The fraudster tries to build a sense of urgency that you need to move your money. Maybe they say the bank account's been.

[Mark Guapozeski]

Compromised, will never ask you to move your money. If, if we think an account has been compromised, the bank will move your money for you, you know, within the bank.

[Narrator/Reporter]

But he says a lot of people believe they're talking to the bank. The supposed bank worker will tell them to buy gold or cryptocurrency or withdraw cash and hand it over to someone thinking that will keep the money safe.

[Mark Guapozeski]

It's a scam.

[Narrator/Reporter]

Every time another voice Scam is going straight to the customer, pretending to be a loved one in need.

[Mark Guapozeski]

Being in this part of the business I have with my family is essentially a safe word. And we all know if there's ever a situation where somebody is either really in trouble and asking for money, it's consistent. We will ask for the safe word.

[Narrator/Reporter]

Angel could have used that he didn't.

[Darian Woods]

Need one, but I thought the voice clone was actually pretty good. And that's why Ben Coleman from Reality Defender doesn't think that protecting banks is enough. He wants all content online to be vetted for whether it was AI generated from text to voice to video.

[Narrator/Reporter]

Yeah, like celebrity scams are a problem right now. We've recently had revelations of scammers pretending to be pro golfers on Instagram and Facebook.

[Ben Coleman]

I think we're gonna look back and say, I can't believe there was a time when we didn't have automated deepfake detection. Our challenge is just the technology is moving quicker than regulations.

[Narrator/Reporter]

And that's why Ben went to Congress to try to advance regulation. That would mean when you logged into Instagram or got a zoom call or a WhatsApp voice mem, you would be informed about whether it was AI.

[Ben Coleman]

We just gave testimony in Congress and in the Senate about this. We deep faked Senator Blumenthal and Senator Hawley Son. We're going to ask the audience and those on the DS which ones are real and which ones are fake.

[Mark Guapozeski]

Hi, my name is Richard Blumenthal, United States Senator from Connecticut, and I'm a die hard Red Sox fan.

[Darian Woods]

That clip of Richard Blumenthal was AI by the way.

[Narrator/Reporter]

Geez, he needs a safe word. Should be Red Sox.

[Darian Woods]

By the way. We at the Indicator are not immune from AI deep fakers. We've heard reports of scammers pretending to be us at the Indicator, so note that we will always use anpr.org email address. Tomorrow in our Vice series, we bring you an episode on what's supercharging? Data breaches. This episode was Produced by Koopacatz McKim with engineering by Robert Rodriguez. It was fact checked by Cierra Juarez. Cake and Cannon edits the show and the Indicator is a product production of npr.

[Edward Jones Sponsor]

This message comes from Bonterra. Your nonprofit's mission is big, but your team is small. With fundraising software from Bontera's network, for good, you can focus on changing lives. Big impact, less stress. Learn [e-mail] NFG this message comes from Mint Mobile. At Mint Mobile, their favorite word is no. No contracts no monthly bills, no hidden fees. Plans start at $15 a month. Make the [e-mail] Switch that's mintmobile.com Switch Upfront payment of $45 for 3 month 5 gigabyte plan required equivalent to $15 a month New customer offer for first 3 months only, then full price plan options available, taxes and fees extra. See Mint Mobile for details.

[Indeed Sponsor]

This message comes from Warby Parker Prescription eyewear that's expertly crafted and unexpectedly affordable. Glasses designed in house from premium material starting at just $95, including prescription lenses. Stop by a Warby Parker store near you.