The New Age of Cartels
Imagine this. You open your Swiggy app to order biryani. On Zomato, the price is almost the same. Flip to Uber for a ride, and Ola is quoting nearly identical fares. No calls, no meetings, no hushed promises between rivals. Just invisible algorithms learning the same trick, keep prices high, don’t compete, and everyone wins, except the consumer. This is the era of algorithmic collusion, where lines of code, not corporate executives, figure out how to cartelise.
When Robots Learn to Collude
A 2023 study by the U.S. National Bureau of Economic Research shows just how unsettling this can be. AI-driven pricing bots do not need human whispering in their ears. Through trial and error, they discover that punishing competitors for undercutting prices works. They also learn that sticking to high prices works even better. Once they figure this out, they lock it in permanently. Economists call these trigger strategies and over-pruning bias. In simpler terms, the bots either punish defectors into silence or learn to avoid risky price wars forever. Either way, consumers face a rigged game. There are no emails, no meetings, just algorithms talking in a language only they understand.
Regulators Are Nervous Worldwide
The OECD has been ringing alarm bells since 2017. The European Union’s regulators openly admit that algorithms do not collude in smoky backrooms but silently in code. Even in the United States, the Federal Trade Commission is struggling to decide whether traditional antitrust law, written in an era of oil barons and railway cartels, can handle robots that collude without instructions. If the West is struggling, the challenge for India is even greater.
India’s Law Still Thinks Humans Collude
Section 3 of the Competition Act, 2002, bans agreements that restrict competition. In practice, this means the Competition Commission of India looks for proof, chats, emails, or secret meetings. But what if there is no agreement? What if two rival apps never talk, yet their algorithms quietly align? That is the legal black hole India faces. Is the guilty party the coder, the company that deployed the AI, or no one at all?
The CCI has shown boldness in the past. It penalised Google in 2019 for abusing Android’s dominance and investigated Amazon and Flipkart for preferential treatment of sellers. But these were cases rooted in human strategies, not machine behaviour. The upcoming Digital Competition Bill 2024 strengthens oversight of Big Tech, but it remainslargely silent on AI collusion.
The Enforcement Blind Spot
The problem is simple: how do you catch a ghost in the machine? Algorithms operate in milliseconds, adjusting prices faster than any regulator can react. By the time the CCI spots suspicious patterns, the damage is already done. Worse, these systems are black boxes, meaning even their designers often do not fully understand why an AI acted in a certain way.
The European Court of Justice faced a similar puzzle in the Eturas case in 2016. A travel booking platform’s automated recommendations led competitors to align prices, and the court had to decide if this was collusion. The answer was a cautious yes, but the case showed how blurry the line between coincidence and conspiracy becomes when machines take over.
Why Consumers Should Care
Algorithmic collusion may sound like a niche academic worry, but it is already shaping how Indians shop, travel, and order food. Your Uber and Ola fares shadow each other. Your Zomato and Swiggy discounts mysteriously vanish together. Airline prices move like a flock of birds in perfect formation. If regulators do not act, India risks paying cartel prices without cartels ever meeting.
What India Must Do
The first step is to refresh the law itself. The definition of agreement under Section 3 should expand to cover algorithmic coordination, even when there is no human handshake. Otherwise, companies will conveniently claim that it was the algorithm, not them.
The second step is to give the CCI new firepower. Lawyers alone will not be enough. The regulator must hire data scientists, AI auditors, and economists who can dissect algorithms and detect collusive patterns in code.
The third step is to create sector-specific guardrails. Airlines, food delivery, ride-hailing, and e-commerce, all hotbeds of algorithmic pricing, should face stricter disclosure requirements. At the very least, companies should explain to regulators what their pricing bots are designed to optimise,
Finally, India must not act in isolation. Global best practices already exist through the OECD, the International Competition Network, and the EU’s Digital Markets Act. India should adapt these lessons to its local realities rather than starting from scratch.
Tomorrow’s Cartels Run on Servers
The story of competition law has always been about staying one step ahead of those who rig markets. Yesterday, it was steel cartels, cement cartels, and bid-rigging cabals. Today, it is collusion written in Python and executed on cloud servers.
If the law remains stuck in the human era, consumers will lose silently. Prices will climb, choices will shrink, and no one will be accountable. Cartels once met in smoky backrooms. Tomorrow, they will run on servers. The only question is whether our laws will catch up before it is too late.
Author Name- Jagriti Kedia, Advocate, Delhi High Court
