---
type: Glossary Term
title: Self-Preferencing
description: Self-preferencing describes the behavior of a platform that systematically favors its own products or services over equivalent third-party offerings — even when
resource: "https://www.contextstudios.ai/glossary/self-preferencing"
category: economics
language: en
timestamp: "2026-06-28T12:05:42.458Z"
---

# Self-Preferencing

Self-preferencing describes the behavior of a platform that systematically favors its own products or services over equivalent third-party offerings — even when that choice is not the best one for the user. The term comes from competition law, notably the EU's Digital Markets Act, and is increasingly applied to the AI market.

In an AI context, self-preferencing shows up wherever a provider controls both distribution and a model of its own. A development environment, an agent runtime, or a cloud platform routes requests to its in-house model by default, even when an equally good or better third-party model is available. Defaults, pricing, and depth of integration are arranged so that the provider's own model holds a structural advantage.

Unlike classic vendor lock-in, the dependency here does not come from switching costs. It comes from a skewed default at the exact interface where user and model meet. For companies, this matters because a seemingly neutral platform recommendation can in fact be a commercially self-interested one — with direct consequences for the cost, quality, and independence of the AI they run.