Instagram, Threads, and TikTok add user-controlled algorithms

Social media platforms are shifting away from one-size-fits-all feeds toward customizable experiences. By introducing tools that allow users to influence recommendation algorithms, these platforms are using artificial intelligence to let individuals shape what they see. This transition is asserted to help social media companies boost user engagement, as customizing these algorithms allows platforms to display content that users are most likely to consume.

To implement this shift, platforms have launched several dedicated features:

• Threads: On June 16, 2026, the platform launched “Your Algo,” a feature building on its “Dear Algo” tool from February 2026. While “Dear Algo” required public posts to signal preferences, “Your Algo” allows users to privately select topics they want to see more or less of. Users can choose how long these preference requests last, selecting durations of one, three, or seven days. For instance, a user can request to see more sports content and less news for a set period. This temporary setting distinguishes Threads from other platforms, which apply permanent adjustments.
• Instagram: In early June 2026, the platform launched “Your Algorithm,” expanding a tool first introduced for reels in December 2025 to the main feed and explore sections. The feature allows users to view the specific topics shaping their recommendations and adjust them directly within their settings.
• TikTok: The platform features a “Manage Topics” tool launched in 2024, which lets users adjust sliders to control content categories in their feeds. In 2025, TikTok expanded this tool with Smart Keyword Filters, an AI-powered feature that automatically filters out selected keywords along with related synonyms. This allows users to customize preferences for various categories, such as sports, travel, and food.

This shift represents a technical evolution in how platforms manage user feeds. Historically, social media ranking models were built with technology that lacked transparency for users, according to Instagram head Adam Mosseri. However, Mosseri noted that large language models (LLMs)—which are AI systems that process and generate text—can make recommendation systems more understandable by showing why content is displayed and letting users explicitly communicate their preferences.

Why it matters

Social media platforms are moving away from opaque, one-size-fits-all feeds toward customizable experiences, using AI to help users influence recommendations and ultimately increase platform engagement.