Charlene Ashley | Is AI making you pay more?

Imagine two passengers boarding the same flight. Same airline. Same aircraft. Same departure time. Same destination. One paid US$463. The other, US$892. We have accepted this for decades. Perhaps one booked earlier or chose a different fare class. Dynamic pricing has become as much a part of flying as the safety demonstration before take-off. But what if they searched at the same time, for the same seat, on the same flight? What if the difference was not about the flight at all, but what an algorithm believed each passenger was willing to pay? That possibility has caught the attention of the United States Congress.

Delta Air Lines has been testing AI-enhanced pricing technology from Israeli company Fetcherr. In July 2025, Delta President Glen Hauenstein said it covered approximately three per cent of its domestic network, with a goal of reaching 20 per cent. Delta liked the early revenue results — a lot. Senators Ruben Gallego of Arizona, Mark Warner of Virginia and Richard Blumenthal of Connecticut demanded answers, warning about algorithms potentially pushing prices towards each consumer’s personal “pain point”. Delta rejects that characterisation, saying it neither uses nor plans to use customers’ personal information to set individualised fares.

That distinction matters. But Delta may be attracting all the attention while the bigger pricing revolution quietly boards the plane.

AI did not invent differential pricing. Businesses have long perfected it. A hotel room overlooking the ocean costs more than one overlooking the car park. Walgreens acknowledges that prices can vary by location, website and app. Uber raises fares when rider demand outruns available drivers. We accept these differences because we understand the economic logic: different demand, different market, different conditions. Businesses have spent decades asking: What will this market pay? Marketers became sophisticated at answering it, segmenting customers by income, lifestyle, geography, age, purchasing behaviour and psychographics, then positioning products and prices accordingly.

AI changes one word in that equation.

The question is increasingly not what the market will pay. It is what you will pay. And ‘willing to pay’ is not the same as ‘able to pay’.

You may not comfortably afford that handbag. But you have searched for it five times, watched three videos, returned to the product page twice, and abandoned it in your shopping cart. To a sufficiently sophisticated algorithm, those are not random clicks. They are signals of desire.

That is where AI becomes marketing’s dream and, potentially, the consumer’s dilemma.

The Federal Trade Commission found surveillance-pricing intermediaries can draw on granular information: location, browsing and shopping histories, demographics, mouse movements and even products left in an online cart. AI can process those behavioural signals at a scale no pricing department could replicate manually. For businesses, this is enormously attractive.

Marketing has always fought for ‘share of mind’ and ‘share of wallet’. If a customer spends US$1,000 annually in your category but only US$200 with you, the strategic opportunity is not the US$200. It is the other US$800.

AI makes that pursuit dramatically more precise. It can test where interest becomes purchase, where purchase becomes hesitation, and where another dollar causes the customer to walk away. The commercial prize is no longer simply estimating the market’s price ceiling. It is getting progressively better at estimating yours.

There is a business case for celebrating. Better pricing can improve forecasting, reduce wasted inventory and match supply more efficiently with demand. An empty airline seat produces precisely zero dollars once the aircraft door closes. An unsold hotel room tonight cannot be sold tomorrow. Sophisticated pricing could mean lower offers for highly price-sensitive consumers, and higher prices for those signalling greater interest.

For shareholders, this can look remarkably like AI making an old corporate dream come true. But dreams have a habit of looking different from the other side of the transaction.

Consumers generally do not object to different prices. We object to rules we cannot see. If I pay more because I booked tomorrow’s flight tonight, I understand the bargain. If I pay more because an invisible system concludes from my behaviour that I desperately want to travel, the transaction feels fundamentally different.

That creates a strategic risk businesses should not underestimate.

Extracting another US$40 because an algorithm believes a customer will tolerate it may look brilliant on this quarter’s revenue dashboard. Losing that customer’s trust, and years of future purchases, is considerably harder to measure. For marketers, that is customer lifetime value.

AI pricing therefore stops being merely a revenue-management decision. It becomes a governance, brand and customer-value decision.

For centuries, shopping has meant comparing products and prices. Surveillance pricing raises the possibility that consumers may increasingly need to compare algorithms — if they know the algorithms are there, that is.

Businesses should therefore ask more than, “Can AI increase our yield?” Pricing optimisation should not sit solely with revenue or technology teams. It should be treated as a customer-value and governance decision.

Boards and executives should know what customer data informs pricing, what consumers are told, and whether today’s price optimisation could quietly undermine tomorrow’s trust. The issue is not whether companies should maximise returns. Of course they should. The issue is how.

Perhaps we are asking the wrong question.

The controversy is not simply whether the passenger beside you paid less for the same seat. Markets have tolerated different prices. The bigger question is what happens when there is no longer a meaningful single price for that seat at all — only the price calculated for you.

For businesses, that could be the ultimate pricing opportunity. For consumers, it makes transparency considerably more valuable. And for boards and regulators, the challenge is deciding where intelligent revenue optimisation ends and invisible manipulation begins.

AI may finally give businesses something marketers have pursued for generations: the ability to turn individual desire into dollars and cents, instantly, and at scale. So when two passengers board the same flight having paid very different fares, the interesting question may no longer be who found the better deal.

The price may still be right. We may simply never know whether it was right for anyone else.

Dr Charlene Ashley is an international business strategist, organisational behaviour consultant and marketing strategist. Her work spans business strategy, market development, organisational performance and international business across multiple industries and markets. Send feedback to cashley@theconsultancyinc.com