How online retailers are using AI to adjust prices by mining your personal data

If you’re going online to buy some last-minute gifts this holiday season, there’s a chance the price you pay will be influenced by what’s known as “surveillance pricing.” Some retailers are using artificial intelligence to set individualized prices online by sifting through personal data, including age, gender, location and browsing history. Ali Rogin speaks with Jay Stanley at the ACLU for more.

Read the Full Transcript

Notice: Transcripts are machine and human generated and lightly edited for accuracy. They may contain errors.

John Yang:

Today is Super Saturday, the last Saturday before Christmas. It's predicted to be the second busiest shopping day of the year. If you're going online to buy some last minute gifts, there's a chance the price you'll pay will be influenced by what's known as surveillance pricing. That's the practice of some retailers to use the power of AI to sift all sorts of personal data center to set individualized prices online.

Things like your age, gender, geographic location and even browsing history could change the price you pay. Ali Rogan spoke with Jay Stanley, senior policy analyst for the Speech Privacy and Technology Project at the ACLU.

Ali Rogin:

Thank you so much for joining us. So what is surveillance pricing and how does it work?

Jay Stanley, Senior Policy Analyst, ACLU:

Surveillance pricing is basically when companies gather a huge amount of data about they're individual customers. And we're living in an era where more data is being collected about us than ever before. Companies take that data and they use it to try to figure out basically how to wring more money out of you when you buy things from them.

What is your pain point? What are you willing to pay, questions like that. And there's a widespread experimentation happening with that kind of pricing in a number of business sectors today.

Ali Rogin:

So what are some examples of that? How are these companies using the massive amounts of data that exist about all of us online to give us these I've heard it described as like personalized pricing?

Jay Stanley:

I mean, right off the bat, there's a lot we don't know because, number one, they're going to claim trade secrets. Number two, they're using AI, which is very opaque in the first place. Even the businesses may not understand what the logic of the AI they're using is.

But this first came into public attention when the president of Delta Airlines, speaking to investors, said that they were planning on using AI to set an increasing proportion of their prices, their airline prices, and that they would use personal information about people to do so. That created a bit of an uproar in Congress and elsewhere, and Delta backed off and said, no, no, we're not going to do this.

But it really put the issue which has been out there for, you know, 10, 15 years on a low key level, really put it into the headlines recently. And so we're seeing attention by Congress, by state legislatures, by the Federal Trade Commission and others. There was a recent report that Instacart has been doing differential pricing. A consumer group did a test and they found that 75 percent of the products instacart were getting set with different prices with different people, with some of the prices 23 percent higher than for some people, than for others.

The Wall Street Journal has reported that retailers like Home Depot were setting different prices for different customers, with customers in affluent areas actually getting charged less. I think that there's a lot we don't know about what's happening, but it's definitely going on.

Ali Rogin:

What is the difference between this type of surveillance pricing and some other examples and types of differential pricing that we've seen out there already, like dynamic pricing?

Jay Stanley:

I think that the core element of surveillance pricing or personalized pricing is that it's based on data they have about you. It's different from, you know, charging more for an umbrella when it's raining. It's different from Uber charging different prices at different times of day depending on the amount of demand for their cars.

Although there have been allegations that Uber, for example, charges more to people whose battery's about to die and things like that would be personalized pricing. Uber has denied that. There can be very complex and sometimes subtle differences among the different types of pricing and that can make it complicated.

But at the end of the day, they have a lot more information about the way the world works, about how different people react to different pricing, different than any individual does. And they are seeking to use that advantage information against individuals.

Ali Rogin:

But all of this is legal, right? And the corollary question to that is what regulations exist or should exist to put some guardrails around this practice?

Jay Stanley:

It's not totally clear that it's legal. It's a bit of a legal gray area. There are, for example, privacy laws, state privacy laws such as California's that may have things to say about this and the way data is used.

There are civil rights laws that have to do with discrimination based on anything based on ethnicity, religion, age, gender, disability status could get a company in hot water if what they're doing starts to implicate those things.

There are state consumer laws, and then there's the Federal Trade Commission, which has a charge to go after deceptive and unfair trade practices. And there could be claims that some of these fall under that. And in fact, the FTC has been investigating this. They did investigation a report last January, and there's a Reuters report that they're looking at it again under the current administration.

So if you're a business that are doing this, you are not necessarily on solid legal ground.

Ali Rogin:

Is there anything that consumers can do to protect themselves from the possibility of being targeted with surveillance pricing?

Jay Stanley:

It's difficult. I mean, at the end of the day, a lot of these are bigger than any one of us. They're social questions, for example. I mean, one of the big problems here is that so much data about us is being collected and flowing to the companies that give them advantages over us.

And that's because Congress hasn't failed to pass an overarching privacy law, as almost all the other sort of advanced industrial countries have done. That would cut off the flow of data at the source that is fueling this kind of strategy.

But certainly you can do your best to hide your identity when you're searching for prices. You can do your best to protect your privacy and be aware of it and check and see whether people you know are getting charged the same price.

Ali Rogin:

Jay Stanley with the ACLU, thank you so much.

Jay Stanley:

Thank you.

Listen to this Segment