Google decided not to seek permission from web publishers for using their data to train AI models in search. This decision was revealed through internal company document during its antitrust trial in Washington.

This document, authored by a Google Search executive, indicated that allowing publishers to opt out would complicate the training of AI models. Instead, Google established a policy that requires all publishers whose content appears in its search results to permit their data to be used for AI features.

This policy was implemented without a public announcement, the documents show. Publishers can only prevent their data from being used in AI by opting out of Google search entirely, a step many cannot take due to Google’s dominant 90% share of the search market.

The ongoing antitrust trial, overseen by Judge Amit Mehta, is evaluating possible remedies for Google’s search monopoly. Closing arguments are scheduled for later this month, with a ruling expected in August.

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ Market power enables data practices that would be impossible in a competitive landscape

Google’s approach to publisher data reveals how market dominance enables practices that would face resistance in competitive environments.

With 89.2% of the general search market and 94.9% on mobile devices, Google can establish take-it-or-leave-it terms for publishers who can’t afford to be excluded from search results 1.

This power imbalance explains why Google could confidently decide to “silently update” their data usage policies “with no public announcement” despite knowing publishers might prefer having opt-out options.

The internal document’s “hard red line” approach demonstrates how monopolistic market positions translate directly into leverage over content creators and their intellectual property.

This pattern of leveraging dominance into adjacent markets (from search to AI) is a key focus in the ongoing antitrust case, with Judge Mehta now considering remedies that specifically address these data advantages 2.

2️⃣ Antitrust remedies directly target the data advantage powering AI features

The DOJ’s proposed remedies reveal how regulators view Google’s data practices as inseparable from its monopolistic position.

Proposed solutions including sharing user data with competitors and potentially selling Chrome address Google’s ability to leverage its search dominance into AI supremacy 1, 3.

These remedies specifically tackle the asymmetric data access that allows Google to train AI models on publisher content without offering meaningful consent options.