Macron defends EU AI rules and vows crackdown on child ‘digital abuse’ | AI (artificial intelligence)
Emmanuel Macron has hit back at US criticism of Europe’s efforts to regulate artificial intelligence and vowed to protect children from “digital abuse” during France’s G7 presidency.
Speaking at the AI Impact summit in Delhi, the French president called for tougher action following global outrage over the use of Elon Musk’s Grok chatbot to create tens of thousands of sexually explicit images of children and growing concerns about the concentration of AI power in a handful of companies.
His words were echoed by UN secretary-general António Guterres, who told delegates including several US tech billionaires that “no child should be the subject of testing for unregulated artificial intelligence”.
“The future of artificial intelligence cannot be decided by a few countries or left to the whims of a few billionaires,” Guterres said. “Artificial intelligence should belong to everyone.”
Bill Gates was scheduled to speak but withdrew at the last minute amid renewed scrutiny over his past connections to convicted child sex offender Jeffrey Epstein.
On Wednesday, White House senior AI adviser Sriram Krishnan renewed his criticism of the Trump administration’s AI regulation, touting the EU’s AI law.
He told delegates he would continue to “scream” against legislation that was “not conducive to an entrepreneur looking to develop innovative technology.”
But Macron told the intergovernmental summit: “Contrary to what some misinformed friends say, Europe is not blindly focused on regulation. Europe is an area of innovation and investment, but it is a safe area, and safe areas win in the long run.”
Research published this month A study by Unicef and Interpol in 11 countries found that at least 1.2 million children last year reported having their images turned into sexually explicit deepfakes. In some countries, one in 25 children (equivalent to one child in every classroom) was affected.
“There is no reason why our children should be exposed online to things that are legally prohibited in the real world,” Macron said. “Our platforms, governments and regulators must work together to make the internet and social media a safe space. That’s why we’re starting a process in France that will ban social networks for children under 15.”
Among the tech executives in attendance was Sam Altman, chief executive of OpenAI, which is facing a legal challenge from the family of Adam Raine, a 16-year-old who took his own life after discussing suicide with ChatGPT.
Dario Amodei, co-chairman of Anthropic, said he was “concerned about the autonomous behavior of AI models, their potential for abuse by individuals and governments, and their potential for economic dislocation.”
Indian Prime Minister Narendra Modi said, “It is imperative that artificial intelligence is safe for children and under family guidance,” and compared the emergence of artificial intelligence to the discovery of fire and described it as “a profound transformation in human history.”
India is trying to position itself as the world’s third AI power after the US and China; Google this week announced it would invest $15 billion in data centers and undersea cables connecting India to the United States and other countries.
Modi said there should be “set levels of authenticity for content in the digital world… people should know what is authentic and what is created by AI.”
The interventions come at a time of growing public concern about the societal risks of artificial intelligence, as the most advanced models remain largely controlled by about four US companies and a handful of Chinese rivals.
Modi has put forward an alternative vision, using India’s 1.4 billion population as a major growth market for technology firms.
He said: “We must prevent an AI monopoly. Many countries view AI as a strategic asset and therefore it is developed in secrecy and its availability carefully managed.”
“But our nation India has a different perspective. Like Ben, we believe that technology can truly benefit the world only when it is shared and open source code is available.”
His comments appear to be aimed at the United States, where leading AI models are not open source and cannot be used or adapted without permission. In contrast, leading Chinese systems such as DeepSeek and Qwen are generally open source.
