Every time tensions flare somewhere in the world, gasoline prices seem to jump overnight. Drivers expect it. The news blames geopolitics, oil traders blame uncertainty, and politicians blame each other. But here's the question almost nobody is asking: If computers can raise prices within hours, why do they suddenly become so patient when it's time to lower them?

Americans have lived with this frustration for decades. The price of crude oil climbs, and gas stations respond almost immediately. Crude oil falls sharply, and suddenly we're told to be patient. Refiners need time. Distributors need time. Retailers need time. Somehow, that urgency only seems to work in one direction.

Regulators have already gone after algorithmic pricing in apartment rentals, and they're looking at hotel rooms, airline tickets, and online retail.

Now a new California lawsuit and a federal push to investigate gasoline pricing suggest there may be another piece of the story that deserves far more attention. It isn't simply about oil markets anymore. It's about artificial intelligence, algorithms, and whether software designed to maximize profits is quietly changing how fuel prices are set across America.

If that sounds like something out of a science fiction movie, think again.

Kalibrating the market?

Kalibrate is a real pricing platform. The company markets its software as an advanced pricing solution that analyzes competitor prices, wholesale costs, local demand, traffic patterns, and countless other variables before recommending the "optimal" price at the pump. By the company's own marketing, it serves many of America's largest fuel retailers and convenience store chains. Retailers use it because it promises to increase profit margins while staying competitive.

There is nothing inherently illegal about any of this. Every major industry now runs on data analytics.

The concern begins when pricing software stops simply reacting to the market and starts shaping it.

On June 22, three California drivers filed a federal class-action lawsuit in Sacramento — and they didn't just sue the software company. They sued the gas stations. Kalibrate is the lead defendant, but so are Marathon, BP, Circle K, 7-Eleven, Speedway, Walmart, Sam's Club, and Albertsons. According to the complaint, Marathon alone runs more than 1,000 ARCO stations in California and has been letting Kalibrate set prices at them since 2020. Circle K, plaintiffs claim, has more than 400 stations on the software. Albertsons, they allege, has been using it since at least 2009.

The complaint alleges that Kalibrate allowed competing retailers to share competitively sensitive information and receive pricing recommendations that discouraged aggressive competition. It describes a "restoration" feature that plaintiffs say lets nearly all the stations in a market raise prices at the same time.

It also quotes Kalibrate's marketing, which according to the complaint tells operators that even in the face of "falling oil prices ... it's critical to avoid a race to the bottom," and warns that cutting your price to win customers "could be making a change that triggers a downward spiral." The plaintiffs call the platform the "central nervous system for a conspiracy to extinguish retail price competition among gas stations."

Price pumping

What does that cost you? Research cited in the complaint found that stations switching to this kind of software raise prices by about 6 cents a gallon on average — and by as much as 30 cents where most of the stations in an area are running it. Plaintiffs point to a real-world example too: They allege that when one California Albertsons turned Kalibrate on, its pump price climbed 3 to 4 cents within days. That sounds small. It isn't. By the complaint's math, a single penny on the statewide average drains $134 million a year from California drivers' wallets.

Kalibrate says it disagrees with the allegations, calls its technology lawful, and intends to defend itself. The retail chains have not yet answered the complaint. No court has ruled on any of it.

But what happens when thousands of competing businesses begin relying on the same algorithm to determine prices?

Price fixing has been illegal in California for more than a century, and the plaintiffs are suing under that old law. What's new is a statute that took effect on January 1 — AB 325, the Preventing Algorithmic Collusion Act — which says plainly that you cannot escape a price-fixing charge by routing the conspiracy through software. Using pricing software is still perfectly legal, but using it to coordinate with your competitors is not.

AB 325 makes it unlawful to use or distribute a "common pricing algorithm" — software that uses competitor data to recommend, align, or stabilize prices — as part of an agreement to restrain trade. Whatever you think of Sacramento, it closed that loophole first, and this case is the first real test of it.

The A-word

Washington is applying pressure of its own — though it is worth being precise about what kind.

On July 3, the Department of Justice and the Federal Trade Commission sent every state attorney general a letter urging them to investigate whether antitrust violations or price gouging are keeping gas prices artificially high. "Recent volatility in crude oil prices does not suspend either the antitrust laws or state consumer protection laws," they wrote, "and it does not authorize companies to manipulate retail prices or collude with their competitors." The letter followed President Trump's complaint, posted to Truth Social on June 23, that falling crude prices weren't reaching drivers.

But read that letter closely and you'll notice something. It never mentions algorithms. Not once. The federal government is going after gas prices with the same tools it has always used, while the argument about the software is being made by three drivers and their lawyers in a Sacramento courtroom. Nobody in Washington has said the word yet. And whether any of these investigations turns up illegal conduct remains to be seen.

Anyone who has driven for more than a few years knows the pattern. Prices spike within days of a geopolitical event, then drift down at a painfully slow pace. Economists even have a name for it: the "rockets and feathers" effect, and they have studied it for decades. Researchers point to several reasons, including inventory replacement costs, consumer behavior, and local competition. None of those explanations necessarily involve illegal activity.

RELATED: License plate cameras will soon track phones, wearables, infotainment, and even your pets

Rick Kern/Getty Images

Dirty work

But artificial intelligence introduces an entirely new variable.

Unlike traditional pricing models, today's software can monitor competitors continuously, process enormous amounts of market data instantly, and recommend price changes faster than any human pricing manager ever could. If dozens or even hundreds of competing retailers rely on similar recommendations generated from comparable market data, the practical result may be less price competition — without anyone ever picking up the phone to coordinate prices.

That possibility isn't unique to gasoline.

Regulators have already gone after algorithmic pricing in apartment rentals, and they're looking at hotel rooms, airline tickets, and online retail. The Justice Department sued RealPage over the software landlords used to set rents and settled the case last November. The concern in every case is the same: that algorithms may accomplish indirectly what competitors have long been prohibited from doing directly.

It's worth being precise about what that settlement did and didn't say. RealPage paid no penalty, and the government made no finding that it broke the law. What the DOJ objected to was the use of nonpublic information from competing landlords — not the software itself. Using an algorithm to price your product isn't illegal. Feeding it your competitors' private numbers is where the trouble starts. That distinction is going to decide the gas station case too.

Technology moves faster than regulation.

Thin margins

This debate also exposes another misconception. When Americans get angry about gas prices, they aim that anger at the oil companies. In reality, what you pay at the pump includes crude oil costs, refining expenses, transportation, taxes, distribution, and retail pricing. Gas stations generally operate on thin per-gallon margins — the National Association of Convenience Stores puts the net at roughly a dime a gallon once credit card fees and operating costs come out — while state taxes and regulatory costs can dramatically affect what you pay locally, particularly in a state like California.

If gasoline prices are rising because of global supply disruptions, consumers may not like it, but they can understand it. Markets move. Wars affect energy. Hurricanes interrupt refining.

But if pricing software is reducing competition by encouraging retailers to move together instead of competing aggressively for customers, consumers deserve answers.

Artificial intelligence is quietly becoming the invisible middleman in countless financial decisions Americans make every day — insurance rates, airline tickets, hotel rooms, online prices, and now what you pay every time you pull up to the pump.

Most consumers never know an algorithm was involved; they simply assume that's what the market decided.

Algorithms don't care whether you're commuting to work, driving your kids to school, or trying to keep your small business afloat. They don't understand household budgets or family vacations. They optimize. That's what they were built to do.

The question was never whether artificial intelligence can set prices more efficiently. It's whether we've quietly allowed machines to redefine what competition means.

Because if software can determine the price of something as essential as gasoline today, what will it be deciding tomorrow?