Social Media: How Algorithms Shape What We See Online

Wednesday 09 September 2026 - 16:24

Behind every video suggested on TikTok, every post displayed on YouTube, or every piece of content recommended on social media lies an automated selection system. Designed to personalize user experiences and boost engagement, these algorithms are now at the heart of a political and scientific debate over their influence and potentially addictive effects.

Content Selected Based on Each User

An algorithm can be defined as a set of rules that allows a computer system to process information and make certain decisions. On social media, it serves to determine the content likely to interest each user.

To establish its recommendations, the system analyzes a multitude of signals: the posts viewed, the videos watched, the content liked or commented on, as well as certain information related to the user's profile.

The goal is then to predict what might provoke a new interaction. A funny video, a topic that aligns with the user's interests, or even content that might elicit a negative reaction can thus be highlighted.

A Strategic Technology for Platforms

Major tech platforms view their recommendation systems as essential elements of their business model. They allow for the real-time selection of a considerable amount of content and tailor the news feed to each individual.

The precise workings of these systems remain generally difficult to ascertain. Companies share little about their internal mechanisms, which also constitute a competitive advantage.

However, this opacity fuels questions from authorities and researchers regarding how recommendations are designed and their consequences.

Why Is It So Hard to Close the App?

The functioning of social media does not rely solely on content personalization. Several features are also designed to encourage repeated visits to the platforms.

The endless scrolling, which allows users to consume posts without true interruption, is one example. Notifications, unpredictable rewards, and constantly refreshed content also contribute to maintaining attention.

The phenomenon is familiar: an initial connection intended for a few minutes can easily extend much longer.

The Role of the Reward System

Researchers are particularly interested in how these mechanisms can affect the brain circuits associated with anticipation and reward.

According to Mark Williams, emeritus professor of cognitive neuroscience at Macquarie University in Sydney, certain features can push users to stay connected longer than they would like, with possible consequences for sleep and well-being.

The repetition of unpredictable rewards can also sustain a cycle of anticipation and search. Quynh Hoang, a lecturer at the University of Leicester, notably highlights the potential role of this unpredictability in stimulating the dopaminergic system.

Australia Aims to Give Users More Control

These questions now extend beyond the scientific community. The Australian Labor government is preparing a new bill focused on online safety.

The text is expected to propose that users be able to disable personalized algorithmic recommendations. The goal would be to give them more control over how content is presented to them.

This initiative comes as authorities increasingly seek to regulate the practices of major digital platforms, particularly concerning minors and the protection of user well-being.

Towards a New Relationship with Algorithms?

The debate is no longer just about whether social media can effectively personalize content. It now revolves around users' ability to choose how this personalization occurs.

While algorithms enable navigation through an ever-expanding digital universe, their ability to anticipate behaviors and capture attention raises questions about freedom of choice, platform transparency, and user protection.

The next step could thus be to provide users not only with personalized content but also with greater control over how this content is selected.