Children are not test subjects: Joint Statement reaffirms children’s rights in the AI era
UN agencies and international organisations have come together to stress that states and tech companies must protect and respect children’s rights in the context of AI by design and by default.
In September 2024, the UN Secretary-General’s High-level Advisory Body on AI concluded that “Children should not be used as subjects for AI experimentation”. Children are not test subjects. Yet untested, unregulated AI systems are being rushed to market, turning children into guinea pigs in the global AI race. 5Rights has come together with UN agencies and international organisations to reaffirm that AI systems must be designed to respect children’s rights by default.
Following a year of engagement led by 5Rights – from co-hosting the conference on Children & AI in Tbilisi to driving joint submissions endorsed by leading organisations worldwide – the Joint Statement on Artificial Intelligence and the Rights of the Child sets out concrete expectations for states and tech companies. Endorsed by 13 UN agencies and international organisations, including the Office of the United Nations High Commissioner for Human Rights, UNESCO, UNICEF, and the UN Committee on the Rights of the Child, the Joint Statement makes clear that AI systems which are not rigorously tested before deployment to be safe or designed to respect children’s rights must never interact with or impact children.
This roadmap is grounded in well-established rights and principles. It builds on UNCRC General comment No. 25 and echoes the standards set out in the 5Rights Children & AI Design Code.
Welcoming the Joint Statement, Sophie Kiladze, Chair of the UN Committee on the Rights of the Child, said:
“We have a duty to ensure that AI becomes a tool for good, one that supports children’s rights, fosters their growth, and protects their future. The true measure of progress is not just technological advancement, but the well-being and rights of every child.”
Clear responsibilities for governments and businesses
Countering tech industry narratives that portray AI as too novel to govern, the Joint Statement echoes existing responsibilities. It reasserts that states must protect children against right violations by third parties, including business enterprises, and it holds tech companies accountable to conduct regular child rights impact assessments to proactively identify and mitigate negative impacts upstream.
The Joint Statement mandates that AI systems likely to impact children must embed privacy- and safety-by-design, and AI-driven recommender systems be designed to not serve children harmful content and to prevent unhealthy emotional attachment.
Speaking at the Joint Statement’s signing ceremony, 5Rights Honorary President Baroness Kidron said:
“Untested, unregulated AI systems are being rushed to market, making children guinea pigs in the global AI race. Echoing the Children & AI Design Code, the Joint Statement provides a clear framework to hold tech companies accountable for ensuring that AI systems impacting children are rights-respecting and age-appropriate.”
As AI systems increasingly shape children’s education, play, relationships, and access to information, the Joint Statement sends an uncompromising message that commercial profit can never come at the expense of children’s rights.
With 2026 marking the fifth anniversary of UNCRC General comment No. 25, the Joint Statement provides state and tech companies with the framework to honour their obligations and responsibilities under the UN Convention on the Rights of the Child – obligations that are clear, established, and can no longer be ignored.