Children, the internet and AI - risks and advice

Today, not only adults, but also children, have AI as part of their everyday digital lives. Children start to build their digital identity at a young age. This raises many questions to consider, such as what traps they can fall into, how to prevent them and how to make the internet a safe place for all children.
Smålandsposten interviewed our colleague Margarita Sallinen , information and cyber security specialist at Omegapoint. Link to the article you will find here (paywall). We took the opportunity to delve a little deeper into the topic that, in one way or another, affects most of us.
What are the traps that children can fall into online today?
Children today can fall into several types of online traps. For example, they can be subjected to grooming, where adults manipulate them into sharing sensitive information or pictures. Cyberbullying is another major problem, where mean comments are spread via social media, for example. Cyberbullying also includes exclusion, where children are deliberately left out of common chats, games or groups, which can create strong feelings of loneliness and exclusion. Children can also be tricked by scammers, for example through games where they are tempted to buy virtual items for real money. From a cybersecurity perspective, they are also at risk of encountering disinformation, hate speech or inappropriate content, often on platforms where adults have less control over what is shared.
How can parents prevent this from happening?
One of the most important things parents can do is to build trust and have an ongoing dialog about online traps, without blaming. Children should feel that they can always tell you if something feels strange or wrong. It's also good to set guidelines together on internet use, such as which apps are okay to use and what never to share with anyone online. In addition, you should keep up to date with the latest trends and platforms to better understand children's digital lives.
What are parents not aware of when it comes to their children online?
Many parents think that the dangers are only on the big platforms, such as Instagram or TikTok, but many children today are also on lesser-known apps, gaming platforms and chat forums that parents don't know about. Parents often underestimate how quickly children adopt new technologies and how social connections can shift from real life to a digital environment where adults have no control.
But in addition to being protective, we also need to talk more with children about their own role in being kind and responsible online. As adults, we need to be good role models.
How do we make the internet a safe place for children?
We create safety through a combination of presence, knowledge and technological tools. Feel free to use parental controls and filtering tools as support, but don't forget that the most important protection is the child's own ability to recognize risks. Talking about what is okay to see, say and do online, and explaining why, helps children build their own healthy attitudes. It's also good to set simple rules around usage early on, such as not using the phone during the night, to build healthy digital habits from the start. Safety also means knowing that there is always an adult to turn to, without fear of being punished.
AI and schoolwork, what are the pros and cons? How can children use AI to their advantage?
AI can be a great tool for children, if used in the right way. For example, it can help explain difficult concepts, practice language, or support creative work such as writing or making pictures. But there are also risks. Today's AI tools are still immature and can produce inaccurate or misleading information if you don't ask the right questions or critically review the answers. AI can "hallucinate" and give answers that sound credible but are wrong.
Therefore, it is important that children learn to use AI as a coach for learning, not as a shortcut to get ready-made answers. They need to practice asking follow-up questions, reflecting on whether the answer is indeed logical and correct, and being aware that AI is not always right. Double-checking answers against other sources and thinking critically about sources is an important part of the process. By using AI in a reflective way, through questioning, comparing and analyzing, children strengthen both their critical thinking and their knowledge base and use technology as a tool for real learning, instead of passively receiving information.
Why should children learn about cybersecurity?
Cybersecurity is as important for children as it is for adults. Today, children start building their digital identity very early. By understanding the basics of cybersecurity, such as the importance of cyber hygiene, responsible online behavior, and measures to protect themselves, their friends and family, such as creating strong passwords, recognizing suspicious links and being careful with personal information, children not only strengthen their protection but also their respect for online privacy. Instilling a natural sense of cybersecurity and privacy early on will help them develop safe and respectful digital behavior that they will carry with them throughout their lives.
How can AI make us safer?
AI can increase security both in society and for us as individuals. In cybersecurity, AI is used to detect suspicious activity, stop scams and anticipate threats before they cause harm.
For individuals, AI can help identify phishing, warn of fake websites, manage passwords more securely and detect if personal data is being spread online. AI tools can also recognize deepfakes, spam, and fake contact requests on social media (which can otherwise be difficult to detect with the naked eye). But it is important to remember that while AI is a useful tool, the human factor, such as understanding the risks and using the technology responsibly, is still crucial to our security.
What do children/young people and parents need to consider when it comes to AI in their free time - for example in social media?
Children encounter AI in apps, games, and social media, among others. Here it is important to think critically about sources, many platforms use AI to adapt content to behavior, which can reinforce certain opinions, affect self-esteem or increase screen time. What we see in the feed is rarely a coincidence. Good questions to talk about together are: Why am I seeing this? Who wants to influence me, and why? Is it a real person or a fake person? You can also use multiple AI services to compare answers to the same question, and then discuss whether the answers are similar or not, and if so, what that means.
The self-image of children and young people is also affected by the content they consume. Parents should talk to their children and young people about the accounts they follow, the clips they click on and how this affects how they see themselves. Today, it's not just about comparing themselves to filtered images, there are now AI-generated influencers, where the person, the images and the texts are all created by AI. They may appear to be perfect, but they are completely made up.
How can adults be there for children now that AI is developing so rapidly?
The most important thing is not to understand every technical detail or every AI tool available, but to be curious and present. Again, it's about creating an open climate and encouraging dialog where children dare and want to ask questions, tell us when something feels wrong or share thoughts about AI. It can be about thinking together about what is real and what is fake, why an AI-generated answer is not always correct, or what you think is okay to create and share. Adults can also contribute by strengthening children's critical thinking and curiosity. It's about talking about, and showing, that AI can be a tool for learning, not just for entertainment.
Want to know more about Omegapoint's expertise in cybersecure digitization? Read more about it here
Insights
Read more articles

Omegapoint's new CEO: "We will be even sharper"

City of Malmö shows why cybersecurity must be practiced - not just planned
