New research shows online platforms use manipulative design to influence users towards harmful choices

Very Large Online Platforms (VLOPs)use manipulative design

A May 2025 investigation by EDRi member Bits of Freedom revealed that users of Very Large Online Platforms (VLOPs) are directed towards certain actions that are not in their best interest, but which benefit the platform. These findings point to possible breaches of Article 25 and Article 27 of the Digital Services Act (DSA).

Even though prohibited by laws like the Unfair Commercial Practices Directive (UCPD) and the DSA, online platforms continue to use manipulative design to influence choices of users. The research looked at Snapchat, Facebook, TikTok, Shein, Zalando and Booking.com and found many examples of deceptive patterns and attention capturing damaging patterns. This research will be helpful for regulators by demonstrating concrete forms of manipulative design, allowing for more targeted actions against possible violations by platforms.

Most important findings of deceptive design on major online platforms

Below is a list of the most important examples of manipulative design found in Bits of Freedom’s research:

• Almost every single type of notification is turned on by default on the social media platforms. It takes a lot of effort for users to choose which notification they do and do not want to receive, there is no proper explanation for how to do this, and the amount of choices can be overwhelming. When notifications are turned off for Snapchat, the platform sends a pop-up message every couple of days, requesting the user to allow notifications. This practice is specifically mentioned in Article 25 of the DSA as a way of manipulation.
• Snapchat and Facebook use an attention-grabbing red badge within the platform. Usually these suggest that there has been a new interaction with another user. However, they are now also used to tell the user that there is new content (such as videos or comments, that the user has not interacted with yet).
• In cookie pop-ups, users are often steered toward allowing all cookies. Platforms do this by making the “accept all” button stand out visually, thus giving it more prominence, which is specifically mentioned in Article 25 of the DSA. They also add text that advocates for users to accept all cookies. Privacy concerns get less attention.
• Certain settings are often pre-selected to increase the odds that users unknowingly agree with these settings. This distorts the ability to make a free and informed decision, as mentioned in Article 25 of the DSA. For example, when creating a new Snapchat account users are given suggestions for “friends” that are pre-selected. This increases the risk of getting into contact and sharing content with other users that they do not know and did not intend to share information with. When users add information to their Facebook-profile,  the privacy-setting (which is small and grey, making it less prominent ) is on “public” by default.
• The order of content (including comments) on social media platforms is automatically determined by profiling recommendation systems. User data and interactions are used to predict what type of content the user might want to see. Article 38 from the DSA states that there should be a non-profiling alternative present, and Article 27 states that this should be directly accessible and selectable from the same location where users see the profiling-algorithm ordered content. However, for the social media platforms investigated for this research, these alternatives were not easily found, and Facebook even prevents users from setting this alternative as a default. In fact, in April 2025, Bits of Freedom, together with EDRi and Convocation Design + Research filed a DSA complaint against Meta for exactly this reason.
• Users receive numerous “fake friend notifications” on Snapchat. It appears as if a user gets a message from a friend or a person that they follow, however, these messages are generated by the platform itself. These messages are often about new content that is available, or suggested content.
• Snapchat uses a variety of gamification elements and patterns that also cause social pressure (‘snap streaks’ with friends, the ‘snapscore’, a friendship level, and sharing the user’s location with their friends).
• On Facebook, it is a very complex task to delete your account, which is a specific deceptive pattern mentioned in Article 25 of the DSA.
• Shein clearly uses more deceptive patterns than other investigated e-commerce platforms. One of these is showing users a fake timer that makes it appear as if a discount is running out. Another one is using specific sentences that suggest that a product is almost sold out and that users should buy it quickly.

How can this research be used and ways forward

This research contributes to the growing knowledge of academics, researchers, regulators, policy advisors, designers and digital rights organisations about how design can influence user’s choices. It also raises user’s awareness about how they can be manipulated online, allowing them to navigate more thoughtfully and empowering them to take action themselves, such as by reporting cases of manipulative design they encounter to regulators. The report also serves as a wake-up call for designers and platforms: these practices will not go unnoticed.

Moving forward, Bits of Freedom aims to work directly with users, focusing on the effects of the most alarming patterns found during this research. Their next subject? Snapchat notification. The goal of the follow-up research is to specifically demonstrate the effect that the manipulative design has on the user’s experience and behavior. These results will allow possible DSA complaints and enforcement actions to be be strenghtened.

Contribution by: Chitra Mohanlal, Tech Researcher, EDRi member, Bits of Freedom