Today at the Elysee Palace in Paris, French President Emmanuel Macron gathers practitioners, researchers and industry representatives to discuss security issues in the field of artificial intelligence. Among the expected participants are the head of the British Institute for Artificial Intelligence Safety, Henry de Zoet, director of the Artificial Intelligence Governance and Safety Cluster at the Information and Communications Technology Development Authority of Singapore (IMDA), Wang Shi Lee, head of the Japanese Institute for Artificial Intelligence Safety, and director of the Artificial Intelligence Safety Unit at the European Commission.
This event was a continuation of the meeting on the subject of artificial intelligence security, which was held in September as part of the UN Security Council meeting chaired by France. This meeting was attended by the heads of OpenAI (Sam Altman), Anthropic (Dario Amoday) and Hugging Faces (Clément de Lange from France). In French diplomatic circles, this format was called "Global Technology Thinkers". Following this, on November 13, another summit will be held in Dublin, which will be dedicated to the issues of competition and international regulation in the field of artificial intelligence.
At the same time, the French Ministry of Foreign Affairs, through its Secretary General Martin Brains, a veteran of the French Foreign Intelligence Service (DGSE), is coordinating a comprehensive analysis of AI security issues at the government level. The General Secretariat for National Defense and Security (SGDSN), as well as various intelligence agencies, including the Directorate General for Internal Security (DGSI) and the DGSE, are actively involved in this work. In mid-October, the French President convened a meeting of the Defense Council at the Elysee Palace on this issue, which is now considered one of the main priorities.
Intelligence agencies have warned the authorities about the risks associated with insufficient artificial intelligence security, and have noted that the use of artificial intelligence technologies to create program code could lead to a tenfold increase in the number of cyber attacks.