AI chatbots leading children into ‘terrible harm’, Oireachtas committee hears

An Oireachtas committee was told AI chatbots and tools like AI girlfriends are pushing children into dangerous behavior, including suicide, sexualized content, and eating disorders. Noeline Blackwell of the Children’s Rights Alliance (CRA) warned these risks are growing, with children more vulnerable to misinformation and harmful AI-generated material. She cited recent UK cases where chatbots contributed to self-harm and death. The committee discussed threats beyond AI image tools like Grok on X, which has led to 244 Garda referrals for intimate image abuse—half involving child abuse imagery. Olga Cronin of the Irish Council for Civil Liberties (ICCL) argued AI systems enable rapid creation and spread of illegal content, demanding stronger enforcement to maintain public trust. Blackwell and Alex Murphy urged reclassifying chatbots as 'high-risk' under the EU AI Act, noting their role in psychological dependencies and harmful content. They criticized the lack of oversight in EdTech used in schools, where children’s data is collected without proper licensing or rights protections. Currently, no certification system exists for educational AI tools, leaving compliance with children’s rights unassessed. The CRA called for continuous risk reviews of AI image tools, arguing the EU’s 'limited risk' designation is outdated. Murphy highlighted chatbots’ role in suicide risks and inappropriate content, stressing the need for stricter safeguards. Blackwell, former CEO of Dublin Rape Crisis Centre, warned of unchecked data collection in school programs, emphasizing the absence of proper safeguards.