Meta Brings 'Your Algorithm' Controls to Facebook After Instagram Debut

Meta is expanding its AI-driven 'Your Algorithm' tool to Facebook, letting users adjust what the ranking system shows them.
Meta is rolling out a feature called "Your Algorithm" on Facebook, giving users direct control over how the platform's AI ranks the content they see. The tool, which the company first tested on Instagram, marks one of the company's most concrete attempts to hand ranking decisions to users rather than leaving them entirely to automated systems. Meta announced the expansion without disclosing a specific rollout timeline or the full list of controls users will get.
The mechanics matter because Facebook's feed is driven by machine-learning models that score and order posts, Reels, and recommendations in real time. "Your Algorithm" gives users a way to influence those signals, letting them tune what the system prioritizes instead of only reacting to it through likes and skips. That is a meaningful shift for a platform where the ranking model, not the follow graph, has determined most of what people see for years.
Instagram served as the proving ground for the idea. Meta has been testing algorithm controls there as part of a broader push to address criticism that opaque recommendation systems shape attention without accountability. Regulators in Europe have pressed platforms to offer alternatives to personalized feeds under the Digital Services Act, and a user-facing tuning tool is a more direct answer than burying settings in a menu.
For Meta, the move also carries business weight. Its advertising machine depends on engagement, and giving users control over ranking could, in theory, reduce time spent on the app if people narrow their feeds. The company appears to be betting that transparency and retention are compatible, and that users who feel heard stick around longer than those who feel manipulated.
The announcement lands amid a wider industry pattern of AI systems being opened up to user adjustment. Recommendation engines at TikTok, YouTube, and X all rely on similar model-driven ranking, and each has faced pressure over what those models amplify. Meta's decision to extend the controls to its largest audience makes it the most visible test of whether user-tuned algorithms actually change behavior.
What remains unclear is how much the controls genuinely alter the underlying model versus adjusting surface preferences. Meta has not said whether the feature changes the ranking weights themselves or simply filters outputs after the fact, and that distinction will decide whether "Your Algorithm" is real user agency or a well-labeled dial. Early users on Instagram will be the first signal, and Facebook's much larger base will be the real test.