Sony Files Patent For AI Matchmaking Based On Search History and Ecosystem Data

A new patent from Sony outlines a system that could implement artificial intelligence matchmaking based on a user's search history to pair players on devices like PlayStation. Game developers traditionally utilize various methods to tackle matchmaking problems, relying on skill levels, in-game rank metrics, personal preferences, or combinations of these factors. Sony's recent filing describes an AI-driven matchmaking system that considers user data across the broader Sony ecosystem.

Published on September 10, 2026, the patent details two categories of information analyzed by the AI model for each player. Explicit data includes normal preferences willingly provided to Sony, preferred games and genres, and preferred player types. This information can be gathered prior to engaging with a title or during profile creation. Meanwhile, implicit information is gathered from multiple services across the ecosystem.

According to the patent, implicit inputs may include time data associated with titles in the game environment, player skill levels, shopping cart information, social media activity, and user search history information. Additional embodiments highlight inputs that include biometric data, chat information, and audio data. Matchmaking under this system relies on activities spanning the entire Sony ecosystem rather than behavior restricted to a single specific game.

The patent also describes a feedback-based updating system where players can update their matchmaking profiles, and are prompted to do so at set intervals or as preferences change. Furthermore, a menu would display recommended matches exceeding a specific compatibility threshold, such as 80 percent, 85 percent, or 90 percent compatibility. The system also factors in complementary skills to fill gaps for players, such as matching a strong support player with suitable tank and duelist roles.