Big announcement planet money is publishing a book and, as of today, it is finally available for pre-order at planetmoneybookcom.
The book is called planet money a guide to the economic forces that shape your life.
It has lots of stories, all that can help you use economics to make life a little richer from work and career to retirement and food, to dating and family.
We'll see you next time.
Help us get some strong pre-order sales.
You'll get a free gift and a free month of Planet Money Plus if you aren't already a subscriber.
Go to planetmoneybook.com.
And thank you.
This is Planet Money from NPR.
So, Waylon, the other day I decided to give our colleague Angel Carreras a little test.
Hello, Darian.
Hey.
Sorry, do you have headphones on?
Angel, I need your help with something real quick.
What's up?
I need your help with something real quick.
You sound like an AI.
Would you be able to go out and buy some gift cards for me?
You sound like an AI.
I don't have time today.
It's a surprise for our colleague.
Darian, I'm so sorry.
You're not fooling me.
This is AI.
If you could just go out and buy like $200 in gift cards, that'd be awesome.
Okay, actually, you've convinced me.
Where should I get these?
D-A-I, Darian?
Come on, quick, on your feet, Darian.
Angel, I have to admit something to you.
Too sloppy, Darian.
This is not my real voice.
Too sloppy.
I am a deepfake.
Gosh.
Angel, I got you.
You got nothing.
You got nothing.
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?
No, I know.
And so yeah, depending on the person, depending on the situation, like maybe it had been urgent, like a cousin calling from a hospital, this could have really fooled somebody.
Yeah, a lot of people are falling for these kinds of audio deepfakes.
It's 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.
Businesses are also getting targeted.
So naturally, they're fighting back, using AI to fight AI.
Hello and welcome to Planet Money.
I'm Darian Woods.
And I'm Weyland Wong.
Normally, we are the co-hosts of Planet Money's short daily podcast, The Indicator.
We're over here today to bring you a couple stories on AI and the evolving business of crime from The Indicator's Vice Week series airing this week.
Today defending against AI voice clones, how the fakes can be detected, what banks are doing to fortify themselves and what we can all do to keep ourselves a little safer.
Then after the break, AI market manipulation.
The new breed of trading bots behave differently and collude differently too.
And regulators can't keep up.
Stay tuned.
Banks are a big target of AI voice fraud.
I mean, that's where the money is, literally.
And so if you're a chief information officer at a bank, you are constantly checking your websites and phone line systems for any vulnerabilities.
That's what Mark Kwapiszewski is doing at PNC Bank.
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.
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 banks' voice verification on the phone.
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.
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.
But we didn't have the buzzword of deepfakes.
We didn't have generative AI.
We didn't have ChachiBT 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.
How did you see this coming?
I just assumed if I was a hacker, what would I be doing?
How do I do more hacking?
That's what I ask myself every day, Darian.
Go to Black Hat Hacker, Will and Wong.
Watch out.
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.
We're doing what's called inference, which is looking for different features that probabilistically indicate that AI was used.
An AI voice has a particular harmonic structure that the human ear doesn't hear, but reality defender software can detect.
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.
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?
Yeah, turn it in.com.
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.
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.
Ben thinks banks should just remove seeing a voice as a kind of password entirely.
We asked Mark Kwapiszewski at PNC Bank why it was still using voice ID.
Are there risks to PNC using voice authentication technology?
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 that we're always looking at is 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 accordingly.
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.
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 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.
The fraudster tries to build a sense of urgency that you need to move your money.
Maybe they say the bank account's being compromised.
We'll 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.
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.
It's a scam every time.
Another voice scam is going straight to the customer pretending to be a loved one in need.
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.
Angel could have used that.
I mean, he didn't need one, but you know, 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.
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.
I think we're going to look back and say I can't believe there was a time when we didn't have automated deepfake detection.
Our challenge is just this technology is moving quicker than regulations.
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 memo, you would be informed about whether it was AI.
We just gave testimony in Congress and in the Senate about this.
We deep-faked Senator Blumenthal and Senator Hawley.
We're going to ask the audience and those on the DS which ones are real and which ones are fake.
Hi, my name is Richard Blumenthal, United States Senator from Connecticut, and I'm a diehard Red Sox fan.
That clip of Richard Blumenthal was AI, by the way.
Geez, he needs a safe word.
Should be Red Sox.
And that's not even getting into video deepfakes.
I've seen these clips coming out of Sora 2, AFN.ai's new video generator.
And what can I say?
I'm not going to believe anything I see on the internet anymore.
Even The Indicator itself as a podcast has been dragged into all this.
We've heard that scammers have impersonated The Indicator and other podcasts.
They invite business owners on as guests.
Then in a Zoom meeting, pretend to be podcast staff to interview prep.
And these people have tried to hack them or get sensitive information out of the victims.
So please, if you hear from us, always check for that npr.org email address.
Next up from the Indicators Vice series, the stock market and AI.
Automated trading bots have been around for a long time, but the new bots built with certain types of machine learning can behave differently and can show an inclination toward manipulating the market.
And of course, the tech is moving faster than the regulations.
After the break why it can be so complicated to determine blame when it comes to market manipulation from artificial intelligence.
Adrian Ma picks it up after this.
For almost as long as there have been financial markets, there have been people trying to manipulate those markets.
People who made money off unsuspecting investors by artificially influencing the price of a stock or some other security.
To help us wrap our heads around it, we're going to sort them into two basic buckets.
The first is what you might think of as human-led manipulation.
And to help us explain, we called up Nicole Turner-Lee.
She's director of the Center for Technology Innovation at the Brookings Institution.
When we talk about artificial intelligence today, we're probably talking about it in ways that are attainable for everyday people.
You know, what AI looks like in retail or how AI affects employment.
But when you begin to get into the markets and trading area, AI is so much more opaque.
A couple of years ago, Nicole was asked to research how the increasing adoption of artificial intelligence could shape financial markets.
I have to admit, when I did this project, out of all the projects, this one was the most scary.
Because any AI that goes awry has the potential to shut down the whole market in ways that I think would be unforeseen and consequential.
In her research, Nicole outlined ways AI could be used to mess with markets.
One big one is by spreading misinformation.
And as we all know, information moves markets.
We see this every day in the analog world.
Announcements get made by the president on tariffs.
The market goes awry.
When you have something that's embedded in the technical system.
What's even scarier about this area is you may not know where it emanates from.
Ghost in the machine.
Generative AI makes it super easy for bad actors to manufacture misinformation.
With a few keystrokes, anyone can make a fake news article or a deepfake recording.
And with the help of bots, they can easily spread that misinformation across the internet.
So those are examples of how humans can use AI to manipulate markets.
But what if AI could use AI to manipulate markets?
What?
What we're talking about here is AI-powered trading bots run amok.
Now, trading bots are not an entirely new thing, of course.
For years, hedge funds have been using them to carry out high-frequency trading.
But, as finance professor Ekaterina Svetlova says, those bots still require a lot of human input to tell them how to trade.
In case of high-frequency trading, you give a machine clear rules what to do.
And now we have algorithms which don't receive clear rules from humans.
At the University of Tevente in the Netherlands, Ekaterina studies how technology shapes financial markets.
And she says these new algorithms have given rise to a more intelligent kind of trading bot powered by AI machine learning.
In particular, something called reinforcement learning.
That's where an AI agent is given a goal, like maximize long-term profits.
And without any further instructions from a human, the AI goes to work, figuring out, through trial and error, the best trading strategy to achieve that goal.
And they are not programmed to do this directly.
One way to think about it is that the old trading bots are like Roomba, while the new trading bots are more like R2-D2.
Yeah.
So friendly and helpful.
Maybe not as friendly as R2-D2.
But with that amount of intelligence and autonomy.
Exactly.
Okay.
And yet Katarina acknowledges that it is hard to say right now just how many firms are employing these new kinds of bots.
But she thinks soon financial markets could be filled with them.
And that could cause some problems.
The fundamental models on which those algorithms are based are very similar, and they might react to market signals in a similar way.
Trading bots could start to mimic each other's behavior, unintentionally buying and selling stocks in tandem, causing wild swings in the market.
Or, as a recent study suggests, these AI super traders could even engage in collusion.
Yeah.
Researchers at the University of Pennsylvania ran a simulation to see what would happen if they unleashed a bunch of AI bot traders powered by reinforcement learning into a marketplace.
And what they found was that, instead of trading against each other, like you would expect in a competitive market, these bots started colluding with one another to manipulate the market.
Itak Goldstein is a finance professor who worked on the study.
Collusion.
Here is a situation where they kind of understand over time that they should trade less aggressively against each other because this is going to help others' profits.
And then they kind of start acting like a cartel, right?
What is a cartel?
A cartel is basically when, Instead of each one of them making their own decisions, they're making decisions collectively.
And the surprising thing here is that these bots were not explicitly communicating with one another.
Instead, each of them is using their own machine brains to crunch the data and come up with the best investment strategy.
And it just so happens that the best investment strategy turned out to be for them to act like a price-fixing cartel.
Now, in some of the simulations that Itai ran, there'd be a bot that did not want to cooperate with the others.
And this did not sit well with the cartel bots.
If they see that someone deviated then there will be some punishment that others are going to start trading aggressively to punish the party that deviated.
Vicious.
Yeah, they sent out some digital bots to break their digital bot legs.
Oh, goodness gracious.
I mean, what's wild about this is that if humans engage in this type of financial market collusion and manipulation, they'd be breaking the law.
They might get in trouble with the Securities and Exchange Commission or the Department of Justice.
But the crime of collusion and market manipulation has historically required human intent, you know intent to do the crime.
So here's where things get kind of philosophical, right?
Because can autonomous trading bots even have intent?
And more to the point, who's responsible if a gang of trading bots go on a financial crime spree?
Nicole Turner-Lee of the Brookings Institution says that's a question we don't have good answers to right now.
When things go wrong because AI is not a person, who do you sue, right?
Who holds the liability for these things?
Who is actually liable when you're actually overusing or dependent on AI systems in this marketplace?
And how do you do it in a way where, again, everyday people still have levels of protection?
This is a legal gray area, and Nicole says regulations are probably needed to fix that.
Like all the experts we spoke to, Nicole says AI is not inherently a bad thing for financial markets.
It can be used, for example, as a tool for detecting fraud.
But with regulators still playing catch up.
That means a lot of power rests with the financial firms, who are experimenting with these AIs and still discovering what they can do.
Nicole's advice to them?
Put this on your radar and have literacy around it for your employees and your customers, so that you can ensure if there are bad actors, it's not you.
Are we the baddies?
No.
I don't know.
I mean, are we... You're not the baddie, Adrian.
If we have to ask ourselves that question, what does that say about us?
Remember, the Planet Money book is available for pre-order and we've got some special gifts available when you buy the book early.
And the more books we sell on pre-order, the more momentum we get on book websites and in bookstores.
So please help us kick off the snowball.
Go to planetmoneybook.com.
And thank you.
Today's stories were just two parts of a five-part series The Indicator is airing this week.
Check out the other episodes on The Indicator from Planet Money's podcast feed.
Got a link in the show notes.
Today's show was produced by Cooper Katzma Kim.
It was engineered by Robert Rodriguez and fact-checked by Sierra Juarez.
Kicking Cannon edits The Indicator.
Alex Goldmark is our executive producer.
Thanks for listening.
