Could AI Be the Silent Villain Behind Rising Grocery and Rent Prices?

In 2023, renters in major U.S. cities faced a frustrating pattern. Rent prices across multiple properties, often owned by competing landlords, seemed to rise in unison. It wasn’t just landlords responding to inflation or shifting market trends. Something more elusive was at work.

Unbeknownst to renters, a hidden force was quietly steering these price hikes. AI-driven pricing algorithms, designed to predict optimal prices for landlords and businesses, had found an unexpected loophole. By mimicking their competitors’ pricing behavior, they could keep rates high without any explicit collusion.

But the impact of this goes beyond housing. Grocery prices, food delivery services, and even Uber fares may be influenced by the same invisible force, a phenomenon known as algorithmic correlation.

The Hidden Algorithm That Drives Prices Higher

For decades, businesses have relied on competitive pricing strategies to attract customers. If one grocery store slashed apple prices to $1.99, the rival down the street might counter with a $1.89 discount.

However, recent research from MIT and Northeastern University suggests that AI systems are changing that balance, and not in consumers’ favor.

The study, conducted by Nathanael Jo, Kathleen Creel, Ashia Wilson, and Manish Raghavan, explored the effects of AI models used in personalized pricing strategies. Their findings revealed that businesses adopting homogeneous AI models, algorithms trained on similar data sets or designed with correlated prediction patterns, tend to synchronize their prices.

Even without direct communication, these correlated algorithms behave like copycats. They unintentionally create a synchronized pricing system that reduces competition.

Instead of competing to lower prices, companies’ AI systems naturally mimic one another. This behavior causes discounts to shrink while prices slowly climb.

This isn’t just an abstract concept. Real-world trends are showing signs of this shift.

The Counterintuitive Strategy: Choosing Weaker AI on Purpose

The study also revealed a surprising and counterintuitive strategy. Businesses may intentionally choose weaker AI models to align their pricing with competitors. This decision seems illogical at first, but the study highlights how firms benefit financially from this unusual tactic.

In theory, businesses should always aim for the most accurate model to predict consumer behavior and maximize sales. Accurate models are designed to optimize pricing by understanding customer preferences, spending habits, and market trends.

However, the researchers found that when faced with the option of choosing between a highly accurate yet independent model or a less accurate but correlated model, businesses often chose the latter.

This decision has a clear motivation. Companies using correlated models are better equipped to align their prices with competitors, which can ultimately maximize profits.

Why Weaker AI Models Are Quietly Winning in Pricing

In their study, the researchers simulated various pricing scenarios where businesses could choose between independent and correlated models. Surprisingly, firms using less accurate but correlated models consistently earned higher profits. This happened because correlated models encourage predictable alignment with competitor pricing.

Correlated models, despite being less precise, track competitor pricing patterns more closely. This alignment offers firms a stable way to avoid major pricing risks, especially in industries where consumers are highly price-sensitive.

Here’s how this strategy benefits businesses:

Firms that rely on correlated models are less likely to undercut competitors with aggressive discounts. Precise models may predict lower prices to attract budget-conscious shoppers, but this approach creates uncertainty. Correlated models reduce that risk by mirroring competitors’ pricing behavior. This ensures businesses avoid drastic fluctuations that could hurt their margins.

By mirroring competitor prices, firms also create stable and predictable pricing patterns. This predictability reduces competitive pressure, making it easier for businesses to maintain steady profits.

In industries where discounts heavily influence customer decisions, correlated models reduce the pressure to offer aggressive promotions. As businesses match prices more closely, fewer discounts emerge. This change leads to a quieter form of price control that ultimately benefits the companies while limiting consumer savings.

In one simulated scenario, the researchers discovered that firms facing highly price-sensitive customers consistently chose correlated models over independent models. This occurred 100% of the time, even when the correlated models were significantly less accurate.

This finding reveals a powerful insight. In highly competitive markets, firms may knowingly abandon better AI models if aligning prices with competitors helps them sustain higher profits.

Why Sacrifice Accuracy for Correlation?

Firms may be drawn to correlated models because they provide stability. Even if a stronger model predicts consumer behavior with greater precision, it also creates a higher risk of deviation from competitors’ pricing patterns.

That deviation, while technically optimal, may backfire if rivals fail to follow suit. A firm that lowers prices based on precise AI predictions may lose out if competitors maintain higher prices. As a result, companies often favor correlated models, knowing they will naturally follow the broader pricing trend.

This strategy mirrors a psychological behavior known as risk aversion. People and organizations often prefer predictable outcomes over potentially larger but uncertain gains. For businesses relying on automated pricing algorithms, correlation offers a safer, more predictable approach.

Even if correlated models are weaker in terms of accuracy, they provide a form of “pricing insurance” that allows firms to match competitors without major risks. In a crowded market, this safe and coordinated approach makes correlated models appealing.

How AI Pricing is Already Affecting Grocery Bills and Rent

The dangers of correlated AI pricing are no longer theoretical. Evidence shows that these strategies are already influencing industries that rely heavily on data-driven pricing models.

In 2022, RealPage, a property management software company, faced legal scrutiny for its role in driving rent hikes across multiple U.S. cities. The company’s AI-driven pricing tool recommended rental prices to landlords by analyzing competitor data. Instead of advising landlords to undercut competitors with lower prices, the software encouraged them to raise prices in sync with nearby properties.

This resulted in a wave of rent increases that spread across entire neighborhoods. Even when landlords operated independently, their pricing moved in lockstep. The AI tool created a form of synchronized pricing that mimicked coordinated price-fixing — even though no direct communication between landlords occurred.

The MIT study suggests that this kind of behavior may already be common across industries that rely on data-driven pricing. Grocery chains, rideshare services, and delivery platforms are prime examples. These businesses frequently adopt AI-driven models designed to track competitor behavior and predict pricing trends.

Grocery Prices in the Crosshairs

Supermarkets frequently rely on automated pricing models to optimize product prices based on customer demand, competitor pricing, and seasonal trends. If two competing chains rely on correlated AI models, their prices may start mirroring each other without direct coordination.

For example, if Store A’s algorithm raises milk prices by 10% based on predicted demand, Store B’s correlated model may respond similarly. This happens even without analyzing identical data. The result is a pricing system where both stores maintain higher prices, effectively reducing competition.

This pattern is especially concerning in industries like grocery retail, where small price differences can heavily influence customer behavior. With fewer discounts and synchronized pricing trends, consumers may struggle to find competitive deals.

Food Delivery Platforms and Rideshares

Food delivery services and rideshare apps also rely heavily on algorithmic pricing models. In 2023, consumers noticed a pattern where Uber and Lyft fares often mirrored each other during peak hours. While this pattern could appear deliberate, experts suggest AI models trained on similar data may unintentionally align prices.

Both platforms rely on algorithms that predict demand, estimate travel times, and analyze user behavior. When both systems respond to similar data patterns, they often suggest similar price changes. This unintentional coordination allows rideshare companies to sustain higher fares during busy periods.

The same pattern is emerging in food delivery platforms, where services like DoorDash and Grubhub show parallel price surges. If two competing services rely on similar models, they may unintentionally match delivery fees and service charges. Consumers expecting competitive rates may find that prices have stabilized at higher levels.

The Real Danger: Consumers Lose

The MIT study warns that when businesses prioritize correlated models, consumers face fewer choices and higher prices. Discounts become rare, and the traditional forces of supply and demand become less influential.

This shift challenges the common belief that AI-driven pricing models are designed to improve market efficiency. Instead, the study suggests that these models can quietly reshape markets in ways that undermine consumer interests.

In a marketplace increasingly shaped by algorithms, consumers are left navigating a digital battlefield where competition appears active but is quietly constrained by invisible pricing patterns.

What Can Consumers Do?

While algorithmic correlation is difficult to detect, consumers can take steps to protect themselves. Comparing prices across multiple platforms is crucial. Relying on a single app or website may hide better deals elsewhere.

Watch for sudden, synchronized price surges. If multiple stores, delivery apps, or property managers raise prices at the same time, it could indicate correlated algorithms at work.

Advocate for transparency. Consumer advocacy groups are now calling for stricter oversight of AI-driven pricing models to protect consumers from unfair manipulation.

Can Consumers Fight Back Against AI-Driven Prices?

The MIT research raises a crucial question. If businesses are willing to sacrifice accuracy for profit, and if AI systems are quietly steering markets, how far will this trend go?

As AI-driven pricing tools expand into more industries, the risk of widespread pricing echo chambers continues to grow.

And while businesses quietly profit, consumers may face the consequences. Each synchronized price hike quietly drains their wallets.

Could AI be quietly deciding how much you will pay tomorrow?

Have you noticed suspicious price hikes in your area? Share your experience in the comments below!

Reference: “Homogeneous Algorithms Can Reduce Competition in Personalized Pricing” by Nathanael Jo, Kathleen Creel, Ashia C. Wilson, and Manish Raghavan, 22 March 2024, arXiv preprint.

DOI: arXiv:2503.15634