G7 in Paris: A United Front for AI and Minors' Safety

The digital ministers of G7 countries gathered in Paris under the French presidency to adopt a common roadmap. This meeting highlighted key initiatives for the security of artificial intelligence, open source, and the protection of minors online. France, as the president, played a decisive role in guiding the discussions, particularly with the launch of the "Hiroshima Process."

The Hiroshima Process: A Pillar of AI Governance

The Hiroshima Process for AI was presented as the main framework for enhancing the security of artificial intelligence within the G7. This process aims to encourage companies to publicly declare their approach to governance and risk management related to AI. Under France's leadership, 56 French companies and organizations, ranging from startups to large industrial groups, have joined this initiative. These companies span various sectors, such as health and defense, and their participation underscores France's commitment to promoting its economic fabric while encouraging its G7 partners to follow suit.

Ministers acknowledged the potential dangers of AI, including its possible use for cyberattacks or the design of chemical, biological, or radiological weapons. Rather than stifling the development of AI, they opted for a clear-eyed approach aimed at understanding and anticipating these threats while continuing to deploy AI at scale.

Adoption of Common Terminology for Open Source

The G7 also adopted a common definition of open source for AI to clarify what truly constitutes the openness of AI models. This initiative aims to prevent the misuse of the term for marketing purposes. Concurrently, a free tool developed with the OECD has been made available to help small and medium-sized enterprises assess their capacity to integrate AI and to guide public authorities and investors.

Energy Impact of Digital Technology and Protection of Minors

For the first time, G7 members officially recognized the significant environmental impact of digital technology. With the rise of streaming, cloud computing, AI, and blockchain, energy and natural resource consumption is increasing. In response, an international mapping of best practices has been published, covering aspects such as the sustainable design of digital services, equipment reuse, and measuring environmental impact.

The protection of minors online was another crucial point of the meeting. Ministers highlighted the risks associated with inappropriate content, cyberbullying, recruitment by criminal organizations, and abuses facilitated by generative AI, such as deepfakes. To address these issues, measures such as mandatory age verification, built-in protections from the design stage of services, and a ban on child pornography content were adopted. Platforms are encouraged to collaborate with governments and share their data to better understand and combat these threats. The next meeting is scheduled for 2027, under the presidency of the United States.