�山����

Delta is testing technology that could charge you more (or less) for the same flight based on what it predicts you’re willing to pay. While lawmakers are calling it “pain point pricing” and putting privacy at risk, the airline says it’s not using personal data and that the tool simply helps fine-tune prices the airline would already be adjusting.

Övünç Yılmaz

Experts say Delta’s move marks a major shift and a glimpse of where pricing is headed. To unpack what this means for travelers, �山���� Today spoke with��Övünç Yılmaz, assistant professor of operations and a pricing expert at the Leeds School of Business. He’s spent years studying revenue management for airlines, hotels and events, and he sees Delta’s latest move as a window into the future—one that is both interesting but may raise questions about fairness, transparency and consumer trust.

What exactly is Delta’s new AI-driven pricing, and how does it differ from the pricing strategies airlines already use?

Airlines have been using dynamic pricing for years to adjust ticket prices in real time based on demand and availability. The price you see can depend on a few key factors, like how early you are booking and how many seats are still available. For example, if a flight from Denver to New York is filling up quickly, the price will probably go up. But if there are still lots of empty seats close to departure, the airline might lower the fare to attract more buyers.

What’s new about Delta’s approach is that they’re using AI to take pricing a step further, potentially tailoring fares to individual customers. There are really two sides to what they’re doing. On one side, they’re using AI to assist the work of human pricing analysts. That feels natural to me, and I expect we’ll see more of this as AI can help humans make faster and better decisions.

On the other side, they could show two people different prices for the same flight, even if they’re searching at the same time, based on things like their previous bookings or browsing history. That’s what we call personalized pricing, and it’s a big shift from how airline pricing has traditionally worked.

Personally, I think it’s a bold move, and we’re going to see a lot of debate around it.

Historically, have we ever seen this kind of highly personalized pricing in airlines or other industries?

We’ve seen some forms of personalized pricing before, especially in e-commerce, but this kind of AI-driven personalized pricing is still new when it comes to airlines. One early case that got a lot of attention was Amazon in the early 2000s. People noticed that different customers were seeing different prices for the same DVD, and it caused immediate backlash. Amazon said it was just a pricing test, but they quickly pulled back.

Since then, companies have been more subtle. Instead of clearly showing different prices to different people, they’ve been using targeted promotions. You can see these everywhere: ride-sharing apps offering special deals to select users and online retailers sending personalized discount codes based on browsing history. These are generally accepted because they’re presented as discounts rather than personalized pricing.

The Federal Trade Commission is currently studying these kinds of practices, which it refers to as surveillance pricing. Its initial findings suggest that the practice is already widespread, with data covering industries from grocery to apparel, among many others.

Does AI-driven personalized pricing risk exploiting consumers or pushing prices to their “pain point”? What protections are in place, and are there potential upsides for travelers?

Hypothetically, if AI could perfectly estimate each customer’s true willingness to pay, it could absolutely be used to extract more revenue. For example, if a flight is normally priced at $100 and AI identifies that one customer is willing to pay $200, it might try to charge them that amount. At the same time, another customer who appears more price-sensitive might be offered a lower fare, like $80, to secure the sale.

Of course, accurately predicting how much someone is willing to pay is extremely difficult. But AI can make educated guesses based on browsing data and online behavior. For example, the first customer might have recently purchased a luxury bag, suggesting a higher willingness to pay. Meanwhile, the second customer may have been comparing prices across multiple airline websites, signaling price sensitivity.

Without knowing the exact details of how Delta’s AI-driven pricing works, it’s hard to say how far they’re going in terms of the information they use. After the initial public backlash, Delta stated, “There is no fare product Delta has ever used, is testing or plans to use that targets customers with individualized prices based on personal data.” So, for now, we’ll just have to wait and see how this develops.

What changes do you anticipate in the coming years? Will AI and personalized pricing become even more widespread, or could public backlash or regulatory pressure slow down their adoption?

These days, nearly every click and scroll on the way to a purchase is tracked, so it’s not surprising that companies use that data to inform pricing. Over time, I think people will also get used to it. We’ve already seen this happen with dynamic pricing. What once felt strange or even unfair has gradually become pretty common.

That said, it’s still not well-defined how much data companies should be allowed to use, and what exactly they can use it for. Companies can’t legally discriminate based on protected characteristics, but beyond that, the boundaries are unclear. I think we’ll start seeing more legal and policy discussions around what’s fair, what’s allowed, and where the line should be. But in the meantime, I believe it would benefit everyone if companies are transparent about how they use data in pricing.

Is there anything else consumers should know about these strategies?

Yes—first, there’s Delta’s recent comment about “amazingly favorable unit revenues” from the tests they’ve been running. It might be worth taking a closer look at what’s actually driving those results. Is it simply that AI is doing a better job with dynamic pricing compared to their previous system? Or is some of the lift coming from price personalization? I think that distinction really matters.

Second, consumers tend to get savvier over time, especially if they feel pricing isn't working in their favor. If people start to suspect they’re being charged more based on personal data, they might switch devices, clear cookies or use incognito mode. Some may even move toward companies that feel more transparent. In the long run, trust and clarity can be just as important as short-term revenue, especially in industries like air travel where customer loyalty and lifetime value are critical.