What to Watch: When Pricing Software Becomes Antitrust Evidence

The issue is not whether companies use algorithms. It is whether a shared pricing tool gives plaintiffs a way to turn ordinary pricing decisions into an antitrust record.

Most pricing-software stories begin with a reasonable defense. The tool only recommends. Management still decides. Nobody met with a competitor. Nobody agreed to fix prices. The software just helps the company read demand, inventory, capacity, seasonality, and local conditions.

Often, that will be true.

Pricing software is not illegal because it is software. A company using its own data to price better is not suddenly a cartel because the spreadsheet became more sophisticated. The more interesting question is what happens when competitors in the same market use the same pricing system, especially when that system is fed with current, nonpublic, competitively sensitive information from those competitors.

That is when the pricing tool starts to look different.

The risk is not the algorithm. The risk is the structure around it.

RealPage is the cleanest live example. RealPage sells revenue-management software used by landlords to help price apartment rents. DOJ alleged that RealPage’s software used nonpublic, competitively sensitive information shared by landlords to set rental prices, and in November 2025 DOJ announced a proposed settlement requiring RealPage to end the sharing of competitively sensitive information and alignment of pricing among competitors.

That matters because the theory is no longer stuck at the complaint stage.

Newer developments make the investor point sharper. In January 2026, a Federal Register notice for the proposed LivCor judgment described DOJ’s allegation that LivCor’s agreements with RealPage and other landlords to share information and align pricing violated Section 1 of the Sherman Act. The proposed judgment would bar LivCor from using revenue-management software that relies on competitively sensitive data and from sharing competitively sensitive information with other landlords.

Greystar points the same way. DOJ’s proposed judgment with Greystar, the largest U.S. landlord, would bar Greystar from using revenue-management software that relies on competitively sensitive data and prohibit it from sharing competitively sensitive information with other landlords.

That’s the big shift.

The alleged risk is not just the code. It is the market structure created when competitors use the same tool, feed it sensitive information, and receive pricing recommendations from the same system.

The old picture of price fixing is easy. Competitors meet, exchange numbers, and agree not to undercut each other.

The newer picture is less theatrical.

The meeting room may be the software itself.

What to Watch

Watch whether algorithmic pricing claims keep moving from vendors to the companies using the tools.

A case against a pricing-software company can hurt the vendor through product redesign, compliance costs, customer churn, and lower revenue quality. A case that treats customers as participants in the alleged system is different. That turns a software issue into sector risk.

Private settlements are starting to show why that matters. In May 2026, fourteen apartment owners agreed to pay $218 million in a second batch of private settlements over allegations that they inflated rents through algorithmic pricing software. They admitted no wrongdoing, but the payments show that the economic exposure is not staying neatly inside the software vendor.

Its simply not good enough to ask only whether a company uses pricing software. You have to figure out what the software connects.

A lower-risk tool uses the company’s own data, public information, older aggregated data, and ordinary demand indicators. A higher-risk tool is used by competitors in a concentrated market, receives nonpublic competitor information, recommends prices or price floors, discourages discounting, and leaves behind documents suggesting users understood the tool as a way to avoid undercutting each other.

That distinction is more useful than the phrase “algorithmic pricing.”

The phrase is way too broad.

The Capital Case Read

The mistake is that most mainstream media continue to treat this as an artificial-intelligence story. It is closer to an evidence story.

A company can truthfully say it never called a competitor. It can truthfully say the software only produced recommendations. It can truthfully say the final decision remained with management. Those facts may help. They may even win in some cases.

They however do not erase the core problem if the recommendation was built from sensitive competitor information and pushed competing users toward similar pricing behavior.

That is where the vendor becomes more than a vendor. It becomes the connection point. The subscription becomes the common relationship, and the recommendation becomes the shared language. The adoption data becomes proof of how much the system actually moved the market.

This is where investors should avoid both lazy conclusions. Every algorithm is not a cartel. Every vendor relationship is not harmless either.

The facts that matter are practical: the source of the data, how fresh it is, who else uses the system, and whether the tool merely informs pricing or pushes users toward a specific number. The internal record matters too. If sales decks, training materials, or emails describe the product as a way to reduce concessions, stabilize prices, or avoid undercutting, the software starts looking less like neutral analytics and more like evidence.

Those are not policy questions.

They are discovery questions.

Why the Easy Version Is Too Convenient

The easy version says pricing algorithms are either efficiency or collusion. That is too simplistic.

A single company using better analytics to price its own inventory should not be treated as a cartel. More information can make markets work better. The legal risk comes from the shared architecture around the tool, not from the mere fact that software exists.

The harder cases sit in the middle. The software may have legitimate uses. The vendor may avoid explicit instructions to fix prices. Customers may retain final discretion. Prices may still vary.

The question is whether the system makes independent pricing harder to defend.

That becomes more important in concentrated markets. Fewer competitors, higher adoption, sensitive data, and similar recommendations create a different fact pattern than ordinary analytics.

This is also where customer liability changes the economics. If only the vendor is exposed, the issue may be contained. If customers are named, the risk can spread across landlords, hotels, staffing firms, healthcare companies, ticketing platforms, self-storage operators, travel companies, logistics providers, and digital marketplaces.

The exposure can travel with the tool.

What Would Matter

Watch five things.

First, whether plaintiffs and regulators keep naming users, not just vendors.

Second, whether the tool uses current nonpublic competitor data rather than public, company-specific, or older aggregated data.

Third, whether the product merely informs pricing or actively recommends prices, floors, discount limits, or adherence targets.

Fourth, whether cases survive long enough for discovery. That is when sales scripts, customer emails, product manuals, pricing records, and usage data become available.

Fifth, whether settlements require product redesign. Restrictions on what data vendors can collect, how fresh it can be, whether competitor data can be pooled, and how recommendations can be generated may matter more than the headline fine.

Bottom Line

The market should not punish every company that uses pricing software. That would be too blunt.

The better screen is whether the software turns competitors into inputs for one another’s pricing decisions.

That is where the legal theory becomes economically useful.

For vendors, the issue is revenue quality. If the product’s value depends on fresh competitor data, shared market inputs, or recommendations regulators may force the company to redesign, the multiple deserves more scrutiny.

For customers, the issue is litigation migration. A tool bought for pricing discipline can become evidence that pricing was not truly independent.

These cases are not important because every algorithm is dangerous.

They are important because it shows what investors should look for: common vendor, sensitive competitor data, concentrated market, similar recommendations, customer awareness, and documents that make the software look less like analytics and more like coordination.

Better pricing is not the problem.

Better pricing through a system built on competitors’ data is where the risk starts to become economic.