More than a quarter of primary school-aged children are interacting with AI-enabled chatbots, a new survey by an Irish online safety charity has found.
The research has sparked concern over the risks of unsupervised technology and cast doubt on the efficacy of age verification systems in place to help prevent under-age children from accessing social media services.
The survey, carried out by Cybersafe Kids, also highlighted the negative experiences of younger users online despite a decline in smartphone use among younger age groups.
A quarter of the children surveyed said they had been bothered by something they have experienced online, with exposure to horror, hate and sexual content topping the list of concerns.
READ MORE
The survey, of more than 9,000 children, found 26 per cent of eight to 12-year-olds were using chatbots to find out information and to help with homework, while 36 per cent of children aged between 12 and 15 years of age were also interacting with the technology regularly.
That was a sharp increase from the 5 per cent recorded last year for eight to 12-year-olds and 12 per cent for 12 to 15-year-olds.
Chatbots are becoming increasingly prevalent, with OpenAI’s ChatGPT and Google Gemini integrated into a number of mobile and desktop systems.
[ ‘Children are using AI, whether we like it or not’Opens in new window ]
While most children were using the chatbots to get information, 9 per cent of primary school students and 7 per cent of those in secondary school said they were using the AI technology to produce their school work. Others were seeking advice from the technology.
The report found smart device ownership remained high, with 93 per cent of eight to 12-year-olds owning a smart device, typically a tablet or games console. Smartphone ownership in this age group has fallen to 39 per cent from 49 per cent a year earlier.
Why is the delivery of vital infrastructure so slow in Ireland?
The research also found a significant number of children under the age of 13 – 70 per cent – are on social media and messaging apps rated for 13+ users, indicating age verification was not effective at preventing them from accessing the services.
Primary school-aged children are more likely to have a YouTube account, at 60 per cent, compared to WhatsApp or other social media platforms, while the older age group preferred WhatsApp. Roblox is still the most popular online environment among primary school children, at 63 per cent.
“With few safeguards in place, and chatbot technology increasingly embedded in the apps children use, we are exposing them to misinformation, privacy violations, and harmful advice,” said CyberSafeKids chief executive Alex Cooney.
“We need urgent action – not just in Ireland, but across Europe. This requires strong, co-ordinated regulation at EU level, including effective age verification, the removal of recommender systems, and oversight of AI chatbots, instant messaging, and social gaming platforms.
“These measures must be backed by robust parental engagement and comprehensive digital education,” he said.
