Meta hands Facebook users direct control over feed algorithms

This "Your Algorithm" tool functions as a direct interface for the platform's ranking engine, mirroring features recently introduced on Instagram and Threads. Beyond simple binary feedback, the system may prompt users for specific details, allowing them to describe their content preferences using natural language. Meta claims this approach moves away from static algorithmic models toward a responsive system that interprets the intent behind user input.

Underpinning these updates is an expanded reliance on large language models within Meta’s infrastructure. By integrating LLMs into its ranking architecture, the company aims to move beyond traditional engagement metrics to better grasp the context of what a user wants to see. This shift enables the platform to adapt recommendations within a single browsing session, effectively creating a feedback loop between the user's explicit requests and the content delivered to their feed. Users can access these controls via the three-dot menu on individual posts or through dedicated prompts displayed beneath content.