You pull into a fuel station. The price board reads ₹104.50. You drive two kilometres to the next one. ₹104.50. A third station, across town — ₹104.50. You assume it's a coincidence, or that fuel prices are simply set by the market. But what if the three stations never actually agreed on anything — and their pricing software did the agreeing for them?
That is the uncomfortable idea at the heart of what economists call "tacit collusion" through pricing algorithms. It is not a conspiracy. It is not a cartel in the traditional sense. It is something stranger: independent companies, each using automated pricing tools, arriving at the same inflated prices without ever exchanging a word.
The Amazon Case: When the Algorithm Was Designed to Anticipate Rivals
In its antitrust suit against Amazon, the US Federal Trade Commission described an internal pricing tool named Project Nessie. According to the agency, the system identified products where competitors were likely to follow an Amazon price increase, raised the price, and held it once rivals matched. The FTC alleges the tool generated more than $1 billion in excess profit — and that Amazon paused it during periods of heightened scrutiny, then switched it back on.
Amazon disputes this account and says the tool was discontinued years ago. The case is ongoing, and no final finding has been made. But the allegation describes something deliberate: a company designing a system specifically to anticipate and influence rival behaviour.
The Harder Problem: When Nobody Designed Anything
The Amazon case is the version of this story where intent can be alleged. The harder version is the one nobody designs at all.
In 2017, when automated pricing software became widely available to German gas stations, economists later studied what happened. In markets where the software was adopted, prices rose compared to markets where it wasn't. The stations were not communicating. They were not meeting in back rooms. Each was simply using a tool that adjusted prices based on competitors' prices — and the result was higher prices for everyone.
This is the ghost cartel. No agreement. No conspiracy. Just an outcome that looks exactly like collusion, produced by independent actors using the same kind of software.
Why This Matters More Than Ever
Pricing algorithms are no longer exotic. They are standard tools in e-commerce, airlines, hotels, ride-hailing, and increasingly in retail and fuel. They promise efficiency: adjust prices in real time, respond to demand, stay competitive.
But when multiple companies in the same market use similar algorithms — often from the same vendors — the algorithms can learn to coordinate without being told to. They can find the price point that maximises industry profit rather than consumer welfare. And they can do it faster and more precisely than any human cartel ever could.
The consumer impact is direct: you pay more. The legal impact is murkier: existing competition law was written for a world where collusion required meetings, phone calls, or emails. It struggles with a world where collusion emerges from code.
What Regulators Are Actually Worried About
Competition authorities in the US, EU, UK, and India have all signalled concern about algorithmic pricing. The fear is not that companies are secretly plotting. The fear is that they don't need to.
If two competitors both use a pricing algorithm that reacts to the other's prices, and both algorithms are designed to maximise profit, they may naturally settle on a price above what a competitive market would produce. Neither company may even realise it is happening. The algorithm is doing what it was told — maximise profit — but the market outcome is anti-competitive.
This creates a genuine legal grey zone. Traditional cartel enforcement requires proof of an agreement. But what if there is no agreement — only an emergent pattern?
Confirmed Facts vs What Remains Unclear
Confirmed: The FTC has alleged that Amazon used a pricing tool called Project Nessie to raise prices where rivals were likely to follow. Amazon denies the characterisation and says the tool was discontinued. The German gas station research found higher prices in markets where automated pricing software was adopted.
Unclear: Whether Amazon's tool actually produced the effects alleged — that is for the court to decide. Whether the German findings generalise to other markets and other software. Whether existing competition law can successfully address algorithm-driven tacit collusion without proof of an explicit agreement.
Speculation: Some economists argue that algorithmic collusion is widespread and under-detected. Others argue the evidence is thinner than the theory suggests. Both views are represented in the academic literature.
Why This Is Hard to Fix
Even if regulators wanted to act, the tools are limited. You cannot easily prove that an algorithm "agreed" to anything. You cannot easily show that a company intended to collude when it simply bought a standard pricing tool. And you cannot easily unwind a market outcome that emerged from thousands of independent pricing decisions.
Some proposals exist: require companies to disclose when they use pricing algorithms, audit algorithms for collusive patterns, or create a legal presumption that certain algorithmic outcomes are anti-competitive. But none of these are settled law, and all raise questions about innovation, privacy, and regulatory overreach.
The Wider Pattern: Software That Learns to Compete Less
This is not just about fuel or e-commerce. Any market where competitors use similar pricing software is vulnerable. Airlines, hotels, ride-hailing, and online retail are all candidates. The more concentrated the market, and the more similar the software, the higher the risk.
The uncomfortable implication is that competition itself may be quietly eroding in markets that look competitive on paper. Companies still compete on service, branding, and convenience. But on price — the thing consumers notice most — the algorithms may have stopped competing without anyone noticing.
What This Means for You
If you are a consumer, the practical takeaway is simple: when you see identical prices across competitors, do not assume it is a coincidence. It may be the market working. It may also be the algorithm working — against you.
If you run a business that uses pricing software, the takeaway is more uncomfortable. You may be using a tool that is producing anti-competitive outcomes without your knowledge or intent. That does not automatically make you liable — but it does mean you should understand what your algorithm is actually doing.
If you are a regulator, the takeaway is that the old tools may not be enough. Proving an agreement is hard when the agreement is implicit in the code.
What Happens Next
The Amazon case will continue, and its outcome may clarify how far existing law can stretch. In Europe, the German gas station findings have already prompted debate about whether new rules are needed. In India, the Competition Commission has begun examining algorithmic pricing in digital markets.
What is clear is that the ghost cartel is not going away. As pricing software becomes more sophisticated and more widespread, the risk of unintended collusion will only grow. The question is whether the law, and the companies using these tools, will catch up before the damage is done.
Our Take
The most important thing about the ghost cartel is that it does not require villains. It requires only companies doing what they are told — maximise profit — using tools that are designed to learn from competitors. The result can be collusion without conspirators, higher prices without anyone deciding to raise them, and a legal system that struggles to respond because it was built for a world where collusion left a paper trail.
That does not mean every pricing algorithm is harmful. It means the burden is now on companies and regulators to understand what these systems actually do — not what they were intended to do. Because in the ghost cartel, intent is not the point. Outcome is.
Frequently Asked Questions
What is a pricing algorithm?
A pricing algorithm is software that automatically sets or adjusts prices based on data such as demand, competitor prices, inventory, and time of day. It is widely used in e-commerce, airlines, hotels, and fuel retail.
What is tacit collusion?
Tacit collusion is when competitors end up charging similar, higher prices without any explicit agreement. It can happen when companies independently respond to each other's pricing behaviour — including through algorithms.
Did Amazon actually collude using Project Nessie?
The FTC has alleged that Amazon used Project Nessie to raise prices where rivals were likely to follow. Amazon disputes this and says the tool was discontinued years ago. The case is ongoing and no final ruling has been made.
Can pricing algorithms break competition law without anyone intending to?
That is the central legal question. Traditional competition law requires proof of an agreement. Algorithm-driven tacit collusion may produce anti-competitive outcomes without any agreement — leaving regulators to decide whether existing law is sufficient or new rules are needed.
