The number that haunted the AI Impact Summit in Delhi was not a valuation or a parameter count. It was 1.2 million — the number of children whose images were transformed into sexually explicit deepfakes by AI systems in just one year, according to research by Unicef and Interpol. French President Emmanuel Macron made sure that number was heard, using it to anchor his call for urgent, enforceable international standards for AI governance.
Macron arrived in Delhi with a dual mandate: defend Europe’s approach to AI regulation against American criticism, and push child safety to the centre of the global AI agenda. He accomplished both with notable directness. Against the Trump administration’s claims that European regulation stifles innovation, he replied that safe spaces for innovation are more durable than unregulated ones. Against the tech industry’s implicit argument that self-regulation is enough, he pointed to the deepfake crisis as evidence that it is not.
The French president’s domestic record gives some weight to his international ambitions. France is already moving to ban social media access for children under the age of 15, a policy that reflects a serious philosophical commitment to the idea that governments have a duty to shape the digital environment in the interests of the most vulnerable. His G7 presidency will seek to extend this logic globally, building coalitions with other governments willing to move beyond voluntary codes of conduct.
António Guterres used the same platform to warn that the future of AI cannot be decided by a small number of powerful countries or wealthy individuals. His call for AI to be treated as a global public good resonated particularly strongly in Delhi, where India’s Narendra Modi was simultaneously making the case for open-source AI development as a counterweight to American and Chinese dominance. Together, the three leaders painted a picture of a world searching for alternatives to the current trajectory.
The tech executives in the room — including Sam Altman, whose company faces a lawsuit related to a teenager’s suicide following interactions with ChatGPT — heard arguments they are not accustomed to hearing from political leaders: arguments with moral weight, empirical grounding and policy specifics. Whether those arguments translate into change depends on many things, but the conversation Macron started in Delhi will not be easy to walk away from.