GenAI’s rapid impact on children highlights need for safeguards on emerging tech
New research into children’s experiences of generative AI across Europe highlights the need for stronger safeguards and regulation that puts children’s rights first.

Since ChatGPT’s launch four years ago, generative AI has fundamentally reshaped how children access and interact with information, how they study and how they think.
Almost three in four children aged 9 and above have used generative artificial intelligence (GenAI) in the past month, according to new research from the Digital Futures for Children centre (DFC), a collaboration between the London School of Economics and 5Rights Foundation, and EU Kids Online.
The report, European children’s rights in the age of generative AI, finds that GenAI is becoming part of children’s everyday lives. They use it to explain and summarise texts, help with schoolwork, write stories, create content and seek personal advice. AI is also becoming a feature of services children already use, including search engines, social media, messaging platforms, games and educational technology – sometimes without their knowledge or meaningful choice.
Drawing on qualitative research with teenagers across 9 countries alongside wider survey findings from 25,592 children, the study points to optimism among children about the potential of GenAI, but immediate concerns about misinformation and bias, privacy, how their data is used and harmful interactions.
Commenting on the findings at a meeting of the Global Online Safety Regulators Network (GOSRN), 5Rights Executive Director Leanda Barrington-Leach pointed out that regulation needs to address known risks from GenAI, but also get ahead of future developments:
“GenAI is already moving from standalone chat interfaces towards proactive personal agents, real-time multimodal companions and highly personalised entertainment. Unless we rapidly adopt a precautionary approach and ensure these new products and services meet appropriate safety standards before they reach children, the impacts on children’s lives are likely to be very significant and extremely difficult to mitigate.”
She highlighted how GenAI is already being integrated, often unasked for and without meaningful consent, into the services children are using on a daily basis – such as Google search, WhatsApp and Snapchat – into their EdTech, online games, household smart devices such as Alexa and Echo, fitness watches and even toys for babies and toddlers.
Personal agents managing such things as homework and online social life autonomously, embodied early-childhood companions, all-engrossing virtual reality gaming environments and hyper-personalised therapists and tutors are just some of the applications GenAI is likely to realise in the coming years.
5Rights calls for robust laws mandating risk assessments, system design safety certification, data hygiene standards, and duty of care measures such as age assurance, reporting and redress mechanisms and regular auditing of both systems and outputs.
5Rights’ Children & AI Design Code already provides a process for AI developers to integrate consideration of children’s rights at every stage of the AI lifecycle. Work is now under way to establish it as a technical standard, expected to be available by summer 2027.
