There is a comfortable fiction that governments tell themselves about online child safety: that it is primarily a technological problem to be solved by the companies that understand the technology best. Emmanuel Macron’s speech at the AI Impact Summit in Delhi was a direct challenge to this fiction. Governments, he argued, have outsourced their responsibility for child safety to tech companies and have received inadequate protection in return. It is time for governments to take that responsibility back.
The evidence for his argument is the evidence that Unicef and Interpol published just before the summit: 1.2 million children in 11 countries victimised by AI-generated explicit deepfakes in a single year. One in 25 children in some nations. This is the outcome of a decade of tech industry self-regulation on child safety — a record that does not justify continued deference. Macron’s conclusion is straightforward: if voluntary commitments produce 1.2 million victims a year, they are not sufficient, and governments must step in.
What governments stepping in actually looks like is the substance of Macron’s policy agenda. France is pursuing legislation to ban social media for under-15s — a governmental intervention in a space where platforms have previously claimed primary authority. Through the G7 presidency, Macron is pushing for international standards that would give governments the legal framework to hold platforms accountable for child safety outcomes. He is calling for collaboration between platforms and regulators, but collaboration on terms set by democratic governments, not by commercial interests.
His defence of European regulation against the Trump administration’s critique reflects the same logic. The EU’s AI Act is, in Macron’s telling, an exercise of democratic authority over a domain that has previously been left to market forces. The American argument that this authority stifles innovation assumes that innovation in the absence of accountability is beneficial — an assumption that the child safety crisis comprehensively undermines.
Guterres and Modi gave Macron significant international support, the former on global governance and the latter on child-safe technology design. The growing acknowledgment within the tech industry that external oversight is necessary — reflected in Sam Altman’s call for an international body — suggests that the fiction of self-regulation is losing credibility even among its beneficiaries. Macron’s call to action in Delhi was a demand that governments reclaim their role as the guardians of the public interest. It is a demand that has rarely been more justified.