Stefan Ebenberger is board member of the pan European Internet Service Providers Association (ISPA), and Secretary General of the Austrian ISPA. As part of the European Safer Internet Initiative, in which ISPA is involved, they conduct an annual survey. The last one was on AI chatbots, and 500 people aged between 11 and 17 took part.

---

Your survey shows that AI chatbots are becoming everyday companions for young people. What finding surprised you the most?

Stefan Ebenberger: First, as ISPA, the Austrian Internet Service Providers Association, we are an industry association and have been part of the Safer Internet Initiative for more than 20 years. Our aim is to make our socio-political engagement and our priority of being socially and politically active one of our core pillars, alongside industry lobbying. In that regard, every year, as part of the Safer Internet Initiative, we conduct a survey. The last one was on AI chatbots. 

What surprised us most was the nuanced perspective of young people on AI. They recognise the importance of responsible AI use but also rely heavily on chatbots for accurate answers. At the same time, 59% are concerned that AI may lead to not knowing how to think independently. The study reveals a surprisingly mature perspective of young people and adolescents; they are not rejecting AI but are instead asking for better guidance and support.

This also brings us to the fact that age-appropriate education, media literacy, and digital awareness measures are, of course, of high importance, because for us, as an industry association, we see that AI holds tremendous opportunities for the future and strengthens the development of AI applications in Austria and, on a European scale. It must be one of our priorities, given the European and global competition for educated, highly skilled people and the tremendous opportunities AI and its use hold for economies across Europe. 

Given the social and political importance of this issue and our role as an industry, we must support everyone, especially young people. A study on chatbots, highlighting the surprisingly mature perspective of youths, underscores that we still have considerable responsibilities to assist children and adolescents in their understanding and engagement with AI.

---

Based on your research, what are the biggest opportunities and the biggest risks of young people using AI chatbots in their daily lives?

Stefan Ebenberger: Young people recognise that AI chatbots provide helpful answers to specific questions, offer clear, age-appropriate explanations, help individualise learning support, save time, and are constantly available. These are opportunities, of course, and I don't think they differ between young people and grown-ups.

Regarding risk, young people are quite aware, which I consider a mature perspective. It was somewhat surprising that they recognise the often poor quality of information, since AI chatbots are not search engines, and they understand the risks of relying on AI for cognitive tasks. They also see that AI lacks empathy and understanding, which can lead to dependency and manipulation. Additionally, they are aware of privacy concerns and the possibility of false explanations. 

This shows that young people are conscious of key risks and do not blindly pursue opportunities without considering potential dangers. I believe this awareness is positive, and we should support and encourage it.

---

Many teenagers use AI for emotional support and personal advice. At what point does this become a concern rather than a benefit?

Stefan Ebenberger: What we've seen is that 57% said that AI often gives helpful private advice, although 18% also reported feeling uncomfortable because of an AI answer. It becomes a concern when AI is no longer used as occasional support but is used to replace human relationships and interactions. 

It should not serve as a substitute for personal contact, and if it does, especially among teenagers, it becomes problematic. This is particularly concerning if teens begin sharing private or sensitive information heavily, which worsens the issue.

---

More than half of young people trust AI-generated information, and many rarely verify it. Why do you think they place so much trust in AI? What can be done to encourage them to be more cautious?

Stefan Ebenberger: Data protection is not a major issue for adolescents. As shown by our study, their awareness and concerns about how their data is used, especially when posted in an AI chatbot system, are insufficient and clearly need to be addressed. Children must be educated that everything they post online, particularly in AI chats, isn’t truly private, and they should refrain from sharing any sensitive personal information. 

So supporting and explaining the importance of sensitive information and data privacy to them is crucial. It all begins with adults setting an example. Adults should show interest in what their kids are doing, discuss the chatbot's responses, and talk openly with their children about it.

---

AI is evolving much faster than regulation. While policymakers work to catch up, what can schools do now to help young people use AI safely and critically?

Stefan Ebenberger: AI is evolving rapidly and outpacing regulation, so schools and parents should take proactive steps now. They can focus on building critical thinking, media literacy, and digital awareness in young people, helping them question the plausibility of information and engage in conversations with parents, teachers, or peers appropriate to the child's age. 

It all starts also with teachers. Someone must be responsible for training teachers as well. Our education system often faces many new responsibilities and tasks, which need to be integrated into the system and passed on to teachers. Teachers should be supported in this transition. Additionally, these changes should be reflected in timetables and implemented as early as possible.

In Austria, this has taken some time. At ISPA, we have actively advocated for mandatory digital lessons in schools to raise awareness and improve understanding of digital tools and digitalisation. 

There's a need to enhance this further by integrating AI not just in one weekly lesson but as a core part of digital literacy, used daily alongside other tools.

For example, discussions with 14-year-olds might differ from those with younger children, and understanding when to involve parents or teachers is key. 

---

Do current EU rules, such as the AI Act and the Digital Services Act, go far enough to protect minors, or are there still important gaps?

Stefan Ebenberger: I believe it is crucial to highlight the benefits and opportunities provided by AI, social media, and other digital platforms. While it's essential to safeguard privacy and preserve natural neutrality, these protections must evolve alongside technological progress. The Digital Services Act already encompasses many of these concerns, and the European Commission continuously enforces these regulations. We see ongoing increases in enforcement actions taken by the European Commission.

However, youth protection shouldn't rely only on bans; effective regulation must be combined with educating minors to become critical users. We should integrate digital usage awareness, media literacy, and AI usage into our school system and timetables.

It is also our responsibility as a society, as an industry, to handle these technologies responsible and teach and support especially kids, but also teachers, grown-ups, and senior citizens get a criticalmind and to be educated on how to properly use AI, because we all need to be aware that digitalisation is is the key for our society, and also for our economy, and that's why we should really act and be active on as ISPA.

Related article