Interview with Leali Osmančević, PhD, in Kana

Leali Osmančević, PhD, from the University Department of Communication Studies, spoke with journalist Maja Petranović for the new issue of Kana on social networks, algorithms, and child safety online.

It Is Crucial to Give the Child Tools, Not Just Rules

When an algorithm selects what a child will view for the next half hour, it essentially shapes a narrowed, personalized stream of content that continually reinforces itself because the system learns what captures the child's attention, not what is developmentally beneficial.

Interviewed by: Maja Petranović
 

Social networks have become a fixture of children's daily lives, as well as a source of numerous questions and challenges for their parents. The European Commission's new proposal has reopened the question of how to protect children in the digital world—through bans, surveillance, or education for responsible media use? We discuss this with Leali Osmančević, PhD, Senior Assistant at the Catholic University of Croatia and an expert on children and media.

What Does the EU KIDS Act Bring Us?

Dr. Osmančević, the European Commission recently unveiled a proposal for new rules that would bar children under 13 from accessing social networks, while access for those aged 13 to 15 would be permitted only under parental supervision. What does this proposal specifically entail, and what would it change in the everyday lives of children and their parents?

This is a proposal for a Regulation (known as the EU KIDS Act) presented by the European Commission on September 17, 2026, which is now heading into negotiations in the European Parliament and the Council. The framework of age thresholds is tiered and divided into three groups within the proposal: first, for children up to 3 years of age, there is a general recommendation for a complete ban on screen time; for children aged 3 to 13, the introduction of a total ban on social media access is planned; and for children aged 13 to 15, access would be strictly limited and possible exclusively under systematic parental or guardian supervision. The proposal is backed by broad political consensus that includes a report by a special expert committee comprising over 60 specialists.

What is perhaps even more significant than the age thresholds themselves is that this proposal shifts responsibility onto the social platforms. Until now (and indeed for decades), child safety online has been left to and dependent upon parents and educators. Under this proposal, far greater emphasis is placed on regulating the platforms, requiring them to demonstrate that their services are age-appropriate and safe in terms of content, design, and everything else a child is potentially exposed to in the digital space. This represents a major conceptual shift: accountability now also lies with the creator of the final product the child consumes—that is, with the platforms, which must reconfigure and adapt their algorithms.

It is important to emphasize that this is a proposal: debates and negotiations will now proceed in the European Parliament and the Council of the EU, after which the final text, including the age thresholds themselves, may still be revised. It is also noteworthy that the framework does not apply solely to social media in the narrow sense; it encompasses video-sharing platforms, online games, app stores, and chatbots, showing that the Commission views the entire digital environment surrounding a child as a unified ecosystem rather than isolated platforms.

One of the major issues will be age verification. How can we reliably determine a user's age without compromising the privacy of children and other users? Or will this fall under the jurisdiction of institutions, systems, or the state?

This, I believe, is the most sensitive issue of the entire proposal. The solution proposed by the Commission involves age verification through certified solutions independent of the platforms themselves, such as a free EU age-verification app and, over time, the European Digital Identity Wallet. These solutions reveal only a single data point to the platform: whether the user is above or below the age threshold, without disclosing their identity, name, or identity documents. For existing accounts, indirect signals can also be used, such as account creation dates or credit card details.

A core principle embedded in the text is also crucial: if a platform cannot reliably establish that a user falls within the designated age limits, the default assumption is that the user is a child. Consequently, any doubt resolves in favor of protecting the child, even regarding technical specifications.

In principle, this is a safer approach than, for instance, sending a scanned national ID directly to the platform. However, digital rights experts warn that once age verification infrastructure exists, it can easily expand into full identity verification. Furthermore, because the age verification requirement applies to the creation of every account, rather than targeting minors alone, every adult will in practice have to prove their age in some way to open a profile. Thus, the technology exists and can be relatively sensitive regarding privacy protection, but its actual security will hinge on the details of implementation rather than the concept itself.

The proposal addresses not only user age, but also platform accountability. In recent years, lawsuits have been filed in the United States alleging that major digital platforms deliberately designed their features to encourage children to spend as much time on them as possible. Is it realistic to expect that clear legal frameworks and obligations can be imposed on tech companies specifically to protect children?

The U.S. experience can be instructive in this context. More than a dozen U.S. states, for instance, have sued TikTok, alleging that the platform was designed to be addictive for children and profoundly harmful to their mental health. These lawsuits specifically cite doomscrolling (endless video feeds), push notifications, and beautification filters that establish unattainable standards of appearance. A year earlier, dozens of states sued Meta, claiming the company knowingly built addictive features to keep younger users on its platform—specifically Instagram—for as long as possible.

What is novel in the EU KIDS Act is that these disputed practices are prohibited by law proactively rather than adjudicated retroactively. The proposal demands a ban on addictive design features, recommendation systems based on prior digital activity, excessive multihour scrolling, notifications during nighttime hours, and the ability for strangers to contact minors directly.

In practice, this means children's profiles would, by default, be set to private and visible only to accepted contacts, with geolocation, camera, and microphone access disabled, accompanied by straightforward tools for blocking, reporting content, and filing collective complaints. Artificial intelligence tools must also be disabled by default and may not simulate relationships that could foster emotional dependency. Additionally, features like autoplay for subsequent videos and nighttime push notifications will be disabled by default.

It is realistic to expect parts of this to be enforced, as regulation has proven capable of altering the behavior of large platforms when faced, for example, with substantial financial penalties. The real question is whether enforcement and oversight will be sufficiently rapid and consistent compared to the speed with which platforms adjust their products and services.

How Much Do Social Networks Affect Children?

How much scientific evidence do we currently have that the way children use social media alters their habits, attention spans, concentration, and cognitive development? Are we already looking at changes that could shape entire generations, or is it too early to draw such conclusions?

The evidence is indeed mounting, although caution remains necessary when establishing causality and impact. One of the more significant studies, which tracked nearly 7,000 children aged 9 to 13, found that higher social media use correlated with poorer performance two years later on tests of reading, memory, vocabulary development, and verbal expression—a clear indicator of the connection between screen exposure and effects on the child. A systematic review of 23 studies (BMC Pediatrics, 2025) confirms similar findings: excessive screen time and exposure are linked to impaired attention, memory, and concentration, whereas specific, carefully selected educational content can have a positive effect on the language skills of children and adolescents.

We can conclude that social networks can contribute to certain cognitive and behavioral shifts, alongside a variety of environmental factors surrounding the child. Social media is not solely responsible, but existing research is sufficiently robust and unambiguous to fully warrant stronger, more consistent measures to protect children's safety and introduce regulatory reforms. While it may be premature in some respects to conclude that upcoming generations are being irreversibly shaped, what is verifiable and scientifically grounded is that premature, intensive screen exposure and social media use leave lasting impacts on children.

A Child Is Not Merely a User

A child might spend half an hour on a phone and view dozens or hundreds of pieces of content curated by an algorithm. How much does this algorithmic curation influence what a child thinks, desires, and talks about?

When an algorithm selects what a child will view for the next half hour, it essentially shapes a narrowed, personalized stream of content that continually reinforces itself because the system learns what captures the child's attention, not what is developmentally beneficial. For a child whose identity, interests, and opinions are still forming, this means the world is presented through a filter calibrated for engagement rather than well-being. Consequently, peer groups today often share similar algorithm-shaped "micro-cultures" rather than cultures grounded purely in personal experience or family values. This makes it harder for parents and guardians to track what a child is truly thinking, as the origin of those ideas is not always visible or apparent.

This connects directly to Article 10 of the proposal, which mandates that recommendation engines must prioritize a child's explicitly stated personal preferences, disable recommendations based on implicit behavioral signals by default (e.g., viewing duration, click patterns, comments), and refrain from using data collected outside the service itself. This aims to prevent platforms from inferring what might capture a child's interest.

Let us consider a specific scenario: a child sees a popular influencer promoting a particular product, aesthetic ideal, or lifestyle. How can we teach children to recognize where spontaneous content ends and advertising, manipulation, or pressure to look or live a certain way begins?

Children struggle to distinguish between personal and sponsored or paid content produced by influencers they follow, particularly when that content is intentionally designed to appear authentic despite being commercialized. It is crucial to give the child tools, not just rules—teaching them to ask questions such as: Who profits if I buy this, want this, or start doing this? Did this person really look like this before editing and filters?

It is valuable to identify disclosures such as #ad or #sponsored together with the child, discuss the difference between genuine experience and marketing, and point out production elements (lighting adjustments, editing, retakes, etc.). These strategies apply to other content children encounter as well; developing media and digital literacy from an early age, especially critical thinking and questioning skills, is essential. These skills are best learned consistently and systematically through ongoing conversation, rather than through one-off discussions.

The Burden Does Not Fall Solely on Parents

A parent can set a rule that phones are not used after 9:00 PM, but what if a child encounters content during the day on TikTok, Instagram, or another platform that is distressing, sexualized, mocking, or involves inappropriate contact? Who should respond, and how?

This clearly demonstrates the limitations of relying exclusively on parental supervision. A parent or guardian who sets such a rule cannot prevent what happens to a child at 3:00 PM on the way home from school. Therefore, the solution must be multifaceted. First, the platform must provide functional, child-accessible reporting mechanisms backed by prompt moderation. This is precisely what the EU KIDS regulation seeks to establish legally through the principle of safety by design.

Schools need clear protocols detailing whom a child should turn to in these situations, while parents or guardians must cultivate a relationship of trust where the child knows that sharing a problem will not lead immediately to punishment or device confiscation. That specific fear often prompts children to conceal distressing experiences, thereby compounding the risk.

It is also useful to establish a preventive plan with the child ahead of time: whom to contact and how to preserve evidence (taking a screenshot of unwanted messages or content rather than deleting them). Children need to learn that reporting an issue is a sign of maturity, not a cause for lost trust. Crucially, a child subjected to any form of harassment or abuse is never at fault for what happened.

When should the state, schools, and tech companies assume greater responsibility than parents? In other words, what can and should parents manage on their own, and what cannot be solved without systemic support?

Parents can and should foster open dialogue and trust, establish reasonable usage boundaries, nurture healthy habits regarding device use, and stay informed about the platforms their children use. What parents or guardians cannot do on their own is alter algorithm design, compel platforms to modify accessible content, or guarantee that a child will never encounter harmful material despite active parental controls.

This is why the proposal under discussion is so vital: for the first time, it systematically shifts part of the burden onto institutions and platforms through design regulations, mandatory safety verification, and regulatory oversight, rather than leaving everything to individual parental involvement—which is necessary, but insufficient on its own. Sustained collaboration among all parties invested in children's well-being is always superior in the long run—beginning with families and schools, and extending to broader institutions and the state.

What Can the Church Do?

At the Catholic University of Croatia, your professional focus is within communication studies, specializing in the relationship between children and media. What can CUC specifically do, or what is it already doing, to address these challenges?

As a higher education institution, we can operate on three levels: through research (scientifically tracking how children and youth in Croatia actually use media, rather than relying exclusively on data from other countries); through educating future communication specialists, journalists, and educators who will teach media literacy in schools and through the media; and through public engagement, contributing to public discourse and media platforms.

Furthermore, alongside courses in the University Department of Communication Studies that address these issues and educate students, we offer additional initiatives through the newly established Center for Media Literacy.

In your view, what role can the Church play? Should parishes, religious education, Catholic schools, and family ministry take an active part in children's digital formation, or should this remain primarily within the family and school?

Personally, I believe digital formation should not remain confined to the family and school, as children spend a significant portion of their formative years within faith communities such as parishes. Religious education and Catholic schools offer a safe space to discuss human dignity in the digital realm, the distinction between authentic relationships and screen-mediated ones, and conscience formation regarding what a child posts, shares, or consumes.

In this sense, family ministry can provide parents with what they often lack: a space to connect with other parents about shared challenges, reducing feelings of isolation and the assumption that everyone else has things figured out.

The strength and value of the Church's role lies in complementing and supporting families and schools by building communities and values through which a child learns to discern what is good and what is not—even when the screen is turned off and no one is watching. Through such a comprehensive approach and proactive engagement, the child's well-being can truly remain at the heart of society.