Alleyways for effective international AI governance: Cybersecurity and provenance

As part of the French presidency of the 2026 G7 Summit in Évian, this workshop, organized in partnership with the AI Safety Institute, will provide a platform for multistakeholder dialogue on two critical dimensions of global security impacted by AI acceleration: synthetic content provenance and authentication in the age of mainstream generative AI, and the governance of the cyber capabilities of frontier AI models.

According to new research, a rapidly increasing share of our digital ecosystems is now longer generated by humans but synthetic (at least 35% of all new webpages are partially or entirely AI-generated), generative AI is dramatically lowering the cost of elaborate scams, LLMs are now baked into self-editing malwares (e.g. Promptflux et Promptsteal), and frontier models (such as the recently released Mythos) are now posing national and global cybersecurity risks.

We invite you to join us on June 12, 2026, from 9:00 to 11:30, for two panels dedicated to these rapidly evolving policy challenges for which advancing coordinated international governance initiatives is both crucial and realistic.

Speaker lineup will soon be communicated.