At the heart of this virality phenomenon are platform algorithms designed to maximize user engagement. These systems prioritize posts that receive lots of attention and interactions, inadvertently amplifying falsehoods because sensational or misleading posts typically generate more clicks, shares, and comments. As a result, misinformation can rapidly permeate user feeds, swaying opinions, and sometimes fueling real-world consequences. Finding effective strategies to curb this spread without hampering legitimate engagement is a central challenge facing social media companies and researchers alike.

Enter an innovative concept proposed by scientists at the University of Copenhagen, outlined in a recent article published in the journal npj Complexity. Their approach is rooted in the idea of introducing “digital friction” during the sharing process, a deliberate slowdown or interruption intended to prompt users to pause and reflect before amplifying content. The thinking is simple yet powerful: if sharing becomes less instantaneous and thoughtless, users may reconsider potentially misleading posts before pressing that share button.

Lead researcher Laura Jahn, a PhD student specializing in computational modeling, explains how the idea was operationalized. She and her colleague, Professor Vincent F. Hendricks, developed a computer simulation to model information flow across social networks similar to X, Bluesky, and Mastodon. The model tested the effects of small interruptions—digital frictions—such as a pop-up message appearing before content can be shared. These frictions represent psychological “speed bumps” that momentarily slow users down, allowing them a brief window to reassess whether sharing is the right choice.

Their findings are both encouraging and nuanced. While introducing friction clearly reduces the total volume of shares, it does not guarantee an automatic improvement in the quality or accuracy of the shared content. This subtlety highlights a critical limitation: simply making it harder to share does not mean misinformation will vanish—instead, some users might avoid sharing altogether, while others might still share problematic posts despite the roadblock. Thus, friction alone is insufficient to elevate the overall informational quality spreading through these networks.

To overcome this, the researchers expanded their approach by integrating a learning component into the friction mechanism. Instead of a generic pop-up, their model incorporated brief quizzes or informational prompts designed to educate users on the nature of misinformation, including definitions and the platform’s policies for combating fake news. This educational friction encourages users to engage cognitively with the underlying issues before deciding whether to share content. According to Professor Hendricks, these learning interventions stimulate deeper reflection, leading users to become more discerning about the content they propagate.

Combining friction with learning yielded a significant outcome in their simulations: not only did sharing rates drop, but the average quality of content being shared showed marked improvement. Essentially, the researchers demonstrated that an intelligent gatekeeping step can filter out low-quality or misleading information, while retaining—if not enhancing—the circulation of more reliable posts. This dual effect is critical in safeguarding the informational ecosystems of social media without suppressing healthy discourse and user interaction.

Looking forward, the University of Copenhagen team plans to transition from theoretical simulations to real-world applications by conducting field studies. Such studies will involve collaborations with social media platforms or the use of experimental social networks crafted for research. These real-life tests aim to verify whether the benefits seen in computational models translate into actual behavioral changes and reductions in misinformation dissemination on live platforms.

The researchers express hope that their work will inspire tech companies to innovate beyond traditional content moderation, which often struggles with scale and speed. By implementing thoughtful digital frictions combined with educational prompts, platforms may be able to harness user agency and enhance content quality in a manner that complements automated detection systems. This hybrid strategy could prove pivotal in the ongoing battle against misinformation, empowering users to act as informed gatekeepers within their own online communities.

If collaboration with major platforms is unattainable, the team intends to continue exploring these mechanisms through simulated environments designed for social science research. These controlled settings can provide valuable insights into user behavior under different friction parameters and educational strategies, offering a rich resource for iterative improvement of interventions before broader deployment.

This research emerges from the Center for Information and Bubble Studies at the University of Copenhagen, a hub focused on understanding complex information dynamics and social epistemology. By leveraging computational modeling with psychological insights, this interdisciplinary approach exemplifies how cutting-edge science can tackle societal challenges such as misinformation in the digital age.

In sum, the introduction of small digital frictions—particularly when paired with user education—presents a promising avenue for mitigating the rapid spread of misinformation online. While the challenge is immense and multifaceted, these findings highlight a feasible, user-centered approach that could reshape how social media platforms manage content dissemination in the future, promoting a healthier, more informed online public sphere.

Subject of Research: Computational simulations to reduce misinformation spread on social media through digital friction and learning interventions.

Article Title: A perspective on friction interventions to curb the spread of misinformation

News Publication Date: 3-Nov-2025

Web References:

• DOI: 10.1038/s44260-025-00051-1

References:
University of Copenhagen, Center for Information and Bubble Studies, npj Complexity Journal

Keywords: misinformation, social media, digital friction, computational modeling, behavioral intervention, fake news, information quality, user education, misinformation spread, social networks

At the heart of this research lies a recognition that social media use is not a one-size-fits-all activity. While social networks offer avenues for entertainment, connection, and personal development, they can also lead to issues such as wasted time, negative mood states, and even regret over overuse. The Bristol study, presented at the forthcoming Conference on Human Factors in Computing Systems (CHI ’25) in Yokohama, Japan, highlights how individuals vary significantly in their levels of personal investment when engaging online. Crucially, the research suggests that neither excessive nor negligible involvement yields a positive outcome. Instead, the "Goldilocks zone" of social media use—where personal meaning is balanced with a healthy psychological distance—appears to be the key to maximizing well-being.

Employing a novel person-centered machine learning methodology, researchers categorized users based on their motivations and behaviors, moving beyond simplistic demographics or usage metrics. This analytical approach identified four distinct user archetypes: Socially Steered Users, Automatic Browsers, Deeply Invested Users, and Goldilocks Users. Each group exhibits a unique pattern of interaction, psychological response, and self-regulation challenges, underscoring the imperative for platforms to adopt a more nuanced understanding of their user base.

Socially Steered Users often find their social media behaviors tightly governed by peer expectations and social pressures. Their engagement is frequently driven by a need to conform or maintain social standing, which can heighten stress and reduce authentic interaction. Automatic Browsers, in contrast, are characterized by unintentional, often mindless scrolling. They struggle with a sense of purposelessness, frequently experiencing regret as their usage feels disconnected from goals or meaningful interaction. This group represents a significant challenge in combating the compulsive features of many platforms.

The Deeply Invested Users form another critical segment. These individuals integrate social media deeply into their identity, values, and life goals. Although their usage can be purposeful and intense, it sometimes crosses into overuse, leading to adverse emotional outcomes. Interestingly, the Goldilocks Users emerge as the most balanced group, combining recognition of social media’s value with deliberate detachment. This balanced engagement fosters the lowest levels of regret and appears to buffer against the pitfalls experienced by other groups.

By spotlighting these differentiated user profiles, the study illuminates a path forward for social media design innovation. Rather than applying uniform features and interventions, platforms could develop customized tools that correspond to the distinct regulatory needs of each user type. For instance, automatic browsers might benefit from prompts encouraging reflection and intentional use, whereas socially steered users could have interfaces mitigated to handle social pressure, enabling more authentic interactions without anxiety.

This vision for adaptive social media is rooted in sustainable and user-centric engagement, moving away from algorithms that primarily aim to maximize screen time and ad revenue. Instead, design could be centered on fostering meaningful connection, personal growth, and overall well-being. The research underscores how data-driven, person-centered analyses can reveal latent patterns in user behavior that are invisible when viewing aggregate usage statistics alone, providing a foundation for human-centered digital wellbeing strategies.

The methodology behind these insights involved surveying 500 participants through comprehensive psychological assessments paired with advanced machine learning techniques to cluster users into the meaningful typologies described. This combination of qualitative and quantitative analysis enhances the robustness of the findings, charting a future where AI and behavioral science combine to tailor digital experiences thoughtfully.

Importantly, the implications extend beyond social media. The researchers note similar user segmentation patterns across other digital technologies, including gaming and wellness applications. This convergence suggests that principles of tailored digital self-regulation and engagement could become a new standard across the technology sector, addressing a broad spectrum of challenges related to sustainable user experience design.

Looking ahead, the team aims to explore how real-time identification of user types on social media platforms might be implemented, enabling dynamic adjustment of interfaces and features in response to shifting needs. Such advancements would mark a significant leap in digital health technology, prioritizing user autonomy and psychological resilience over passive consumption models.

In an era where concerns about screen time, digital addiction, and mental health are at the forefront of public discourse, this research offers a hopeful blueprint. By grounding platform design in the nuanced realities of user behavior and psychology, social media has the potential to evolve from a source of stress and distraction into a space of controlled, meaningful, and rewarding interaction.

By tailoring online environments to the varied needs of users, social media can help individuals regain control over their digital lives, promoting intentionality, reducing regret, and ultimately enhancing well-being. This person-centered, machine learning-informed approach heralds a new paradigm in digital experience design, one that acknowledges the complexity of human behavior and embraces personalized technology for the betterment of society.

Subject of Research: People

Article Title: Autonomous Regulation of Social Media Use: Implications for Self-control, Well-Being, and Ux

News Publication Date: 7-May-2025

Web References:
https://dl.acm.org/doi/10.1145/3689044

Image Credits: Dr Feng Feng

Keywords: Applied sciences and engineering