# Algorithmic Pricing Can Break Competition Law: How Growth Companies Avoid a Cartel Interpretation

> Algorithmic pricing can trigger competition law violations when rivals share the same pricing model. Here's how growth companies assess the risk in advance.

- Published: 2026-08-17
- Author: Master Mind
- Canonical: https://aimaster.fi/en/artikkelit/algoritminen-hinnoittelu-voi-rikkoa-kilpailulainsaadantoa-nain-kasvuyritys-valtt

Algorithmic pricing can break competition law even when rival companies have never made a single phone call about prices. The risk emerges when competing companies use the same third-party pricing algorithm or pricing service — the result can be effectively aligned pricing without anyone signing an agreement.

Growth company leadership adopts dynamic or AI-driven pricing because it promises better margins and faster reaction to the market. Few stop to assess what happens when three competitors in the same industry buy their pricing from the same software vendor. Competition law calls this a hub-and-spoke arrangement, and it can meet the legal threshold for a restriction of competition without any direct agreement between the companies.

## What is a hub-and-spoke arrangement in competition law?

A hub-and-spoke arrangement describes a situation where several competing companies (spokes) use the same central actor (hub) — for example a pricing algorithm or service — through which competitively sensitive information or pricing logic flows indirectly between competitors. The outcome can resemble direct price coordination even though the companies never negotiated with each other. Under EU competition law (Article 101 TFEU) and equivalent national rules, both direct agreements and de facto aligned market behavior arising through such an intermediary are prohibited.

## When does algorithmic pricing cross the legal line?

Risk increases when three conditions are met at once: competitors use the same or a similar third-party pricing model, the algorithm ingests other players' pricing or demand data as input, and the outcome is a level of price alignment that cannot be explained by each company's own costs. A pricing model built on a single company's own internal data is a different matter from a shared external model — the difference lies precisely in whether a competitor's information flows through the algorithm into your price.

A practical rule of thumb for growth company leadership: if a pricing tool vendor markets its product as one that "optimizes prices across the industry" or uses anonymized competitor data as an input, get a legal assessment before rolling it out. AI does not remove accountability — the company making the pricing decision is always responsible for the outcome, not the software vendor.

## Why is pricing based on your own data a safer starting point?

Competition law risk drops significantly when pricing logic is built on a company's own data rather than a shared dataset. [Master Layer](https://aimaster.fi/tuotteet/master-layer) is a data foundation layer that securely connects a company's existing systems (CRM, ERP, documents) for AI use — so pricing decisions are not based on a shared third-party input that could contain competitor information.

Once the data foundation is in place, pricing can become a task for a dedicated [Master Mind](https://aimaster.fi/tuotteet/master-mind) agent. Master Mind is a set of AI agents that operates on top of Master Layer's data and runs business processes independently — in this case, calculating price based on your own costs, your own inventory, and your own demand, not a shared industry black box. The distinction sounds technical, but in competition law it is exactly the difference between deciding your own price and being part of an arrangement you would not want to sign your name to.

| Feature | Shared third-party pricing tool | Company-specific AI agent on own data |
| --- | --- | --- |
| Input data | Often shared or anonymized industry/competitor data | Own CRM, ERP, and sales history |
| Competition law risk | Rises if multiple competitors use the same model | Stays low when logic is company-specific |
| Transparency for leadership | Black box, vendor-owned model | Company has direct visibility into decision logic |
| Accountability for outcome | Always rests with the pricing company | Always rests with the pricing company |

## How should a growth company check its own pricing risk now?

Start by mapping where your pricing tool gets its input data — is it your company's own data, or a shared industry model? This mapping is part of a broader AI strategy review that examines all AI-driven processes and their data sources. [Master Plan](https://aimaster.fi/tuotteet/master-plan) is an AI strategy sprint that maps where AI creates the most value for your company — measured in euros — and surfaces exactly these kinds of hidden risks before they become a problem.

The same principle applies more broadly to AI procurement decisions: should the solution be built in-house on your own data, or bought as a shared industry service? This decision is covered in more detail in [Build or Buy? How Growth Companies Get the AI Build-vs-Buy Decision Right](https://aimaster.fi/artikkelit/rakenna-vai-osta-nain-kasvuyritys-tekee-tekoalyratkaisun-build-vs-buy-paatoksen-).

## Questions and answers on algorithmic pricing

## Does AI-driven pricing automatically violate competition law?

No. The risk only arises when competing companies use the same shared algorithm or pricing service in a way that leads to effectively aligned price levels. A company-specific pricing model built on your own data is not, by itself, a competition law problem.

## Who is liable if a pricing algorithm violates competition law?

Liability always rests with the company making the pricing decision, not the software vendor. Using AI does not transfer or remove that responsibility — leadership is accountable for ensuring pricing is based on lawful, independent decision-making.

## Can a small growth company face a competition law investigation over a pricing algorithm?

Yes. Competition law has no size exemption — the rules apply to companies of any size if behavior meets the criteria for prohibited coordination. In practice, risk for smaller companies most often comes from shared, off-the-shelf industry pricing tools.

## How should the competition law risk of pricing be checked in practice?

Determine where your pricing tool's input data comes from, whether your competitors use the same service, and whether the outcome is based on your own cost and demand data. If the answer to any of these is unclear, get an assessment before wider rollout.

Competition law risk is not a reason to avoid AI in pricing — it is a reason to build pricing logic on your own data. Book a [free Master Mind analysis](https://aimaster.fi/analyysi) to review where AI can create value for your company safely, based on your own data.

## Frequently asked questions

### Does AI-driven pricing automatically violate competition law?

No. The risk only arises when competing companies use the same shared algorithm or pricing service in a way that leads to effectively aligned price levels. A company-specific pricing model built on your own data is not, by itself, a competition law problem.

### Who is liable if a pricing algorithm violates competition law?

Liability always rests with the company making the pricing decision, not the software vendor. Using AI does not transfer or remove that responsibility — leadership is accountable for ensuring pricing is based on lawful, independent decision-making.

### Can a small growth company face a competition law investigation over a pricing algorithm?

Yes. Competition law has no size exemption — the rules apply to companies of any size if behavior meets the criteria for prohibited coordination. In practice, risk for smaller companies most often comes from shared, off-the-shelf industry pricing tools.

### How should the competition law risk of pricing be checked in practice?

Determine where your pricing tool's input data comes from, whether your competitors use the same service, and whether the outcome is based on your own cost and demand data. If the answer to any of these is unclear, get an assessment before wider rollout.