Artificial intelligence-generated child sexual abuse material rose from about 5,000 detected cases in 2023 to 1.3 million in 2025, information from figures cited by Pedro Sánchez. The Spanish prime minister pointed to the increase as he called for limits at national and European levels, with global rules as the goal, on technology reaching children faster than safeguards can keep up.
He made the appeal during a high-level dialogue on protecting children's rights in the age of artificial intelligence. The figures came from Sánchez's speech; no separate public report has independently verified the full rise from 5,000 to 1.3 million cases.
Sánchez said companies cannot put commercial interests ahead of children's safety and mental health. Businesses developing these systems must answer for their effects on children, rather than leaving families and society to bear the costs. Those include cyberbullying and hate, as well as sexual crime, privacy loss, anxiety and loneliness.
The coalition for children’s rights and protection in the age of AI was initiated by France, Spain and Kenya on July 7, 2026. By September 22, it had expanded to 23 states and eight United Nations entities, the Élysée Palace said.
The warning comes as children spend large amounts of time online. European minors spend up to six hours a day in front of a screen, and one in three adolescents says they regularly encounter inappropriate content.
Sánchez also referred to the latest OECD PISA report, which found a general decline in educational performance that he linked in part to artificial intelligence use. The cited PISA 2025 findings also connect frequent AI use for schoolwork with lower results than more moderate use. That link is correlational and does not show that AI alone caused the decline.
The concern extends beyond children's habits and school results. Sánchez described a digital environment without rules as a lawless jungle and pointed to the concentration of wealth and influence among a small group of technology oligarchs.
In his account, these companies do more than build tools: they shape values and preferences while controlling systems whose risks are already being acknowledged by people developing advanced artificial intelligence.
Spain is part of the regulatory response. The government has co-led the United Nations Coalition for the Rights and Protection of Children in the Age of Artificial Intelligence.
Sánchez said the coalition was one of the first practical results of the United Nations dialogue on AI governance. That dialogue was approved the previous year, when the International Panel of Experts on Artificial Intelligence was launched. The coalition’s roadmap sets out areas for cooperation, including safety and privacy by design, along with benchmarking and accountability. It also covers AI-generated child sexual abuse material, deepfakes and AI literacy. The United Nations launched the roadmap in New York on September 22, 2026, placing Sánchez’s appeal within an established international work programme.
In Spain, the government has promoted a law that would set 16 as the minimum age for access to social networks, with parental controls activated by default.
The proposal drew partly on France and Australia. It remains a legislative proposal, not a completed rule.
At European level, Sánchez credited the European Commission with securing provisions in the European Union Artificial Intelligence Act. Those provisions prohibit non-consensual sexual deepfakes and AI-generated child sexual abuse material.
Spain has also been among the first countries to develop an age-verification system based on the European Digital Identity Wallet.
The coalition's call for common rules fits into a wider Spanish political argument about how rights are protected and enforced. It comes alongside an earlier citizenship dispute involving suspended instructions and CERA registration.
The subject is different, but the institutional challenge is similar: turning political commitments into rules that can work beyond one government or platform.
Sánchez's position leaves little room for voluntary promises from the technology industry. He is calling for cooperation among governments and international organisations, with companies, educators and families also involved. He insists that artificial intelligence must serve humanity and answer to human needs.
That is the test now facing the coalition: turning concern about rapidly expanding harm into enforceable protections. Without limits and accountability, shared standards could fall short, leaving the technology's promised benefits weaker than the damage unchecked use may impose on children who are least able to defend themselves.
