Here’s a new concept to worry about, if you don’t have enough on your plate — “algorithmic collusion.”

Algorithmic collusion, simply put, is the intentional or unintentional collaboration by companies doing business on the internet to keep prices higher, using the automation of algorithms.

In this, they are behaving like a digital version of a classic cartel, which the dictionary defines as “an association of manufacturers or suppliers with the purpose of maintaining prices at a high level and restricting competition.”

Artificial intelligence, machine learning and deep learning (machine learning on steroids) can be used for purposes that enhance market competition or are anti-competitive, wrote Antonio Capobianco of the Organization for Economic Cooperation and Development in a January paper titled “Digital Cartels and Algorithms.”

Positive examples of commercial algorithms are legion: supply-chain optimization; targeted ads; recommendations; product customization; dynamic pricing; price differentiation, and fraud prevention. Anti-competitive algorithms include bias in favor of incumbents’ products and collusion with competitors.

Algorithms can be used for collusion in setting prices for essentially identical goods, such as gasoline at the pump or airline tickets.

The critical factors are transparency — how easy it is to identify pricing changes by competitors — and frequency of interaction.

Here are three scenarios:

1. The competitors agree to match prices and not lower them below a certain range and embed that in the pricing algorithm.

2. The competitors do not discuss the matter with each other, but write their pricing algorithms to not react to a single downward turn in competitor pricing, only multiple downgrades.

3. They use artificial intelligence, specifically deep learning algorithms, which learn and improve from their own experience. For example, an algorithm instructed to maximize throughput and revenue might often conclude that collaborating with opposing algorithms is the most efficient way to “win.”

Capobianco points out that, unlike traditional cartels, which require a small number of large players in a market to collaborate and control prices, a digital cartel can tacitly set artificially high pricing between many players, as the speed of competitors’ algorithms interacting with each other takes place at the blink of an eye — thousands per second on today’s blazing-fast cloud-based systems.

“If markets are transparent and companies react instantaneously to any deviation, the payment from deviation is zero; collusion can always be sustained as an equilibrium strategy,” Capobianco said.

Once again, this scenario presents a new challenge to global and national regulatory systems.

If ignored, the result will be de facto digital cartels that limit customer choice.

How to regulate these systems is an open question. It will take years to calibrate, design and pass legislation that effectively addresses these complex and nebulous challenges.

This issue of self-learning algorithms coming to conclusions that are opposed to human goals and even ethics will manifest itself more frequently as deep learning becomes more embedded in the global network.

 

Isaac Cheifetz is an executive recruiter and strategic résumé consultant based in the Twin Cities. His website is catalytic1.com.