France’s Take on Artificial Intelligence and on the EU Artificial Intelligence Act Under Discussion

“I would like France to be one of the leaders in artificial intelligence; I would like Europe to be one of the leaders in artificial intelligence.”

This statement by President Emmanuel Macron in 2018 sets out clearly the importance French authorities place on artificial intelligence (AI) and the possibilities it offers. This political statement has been given concrete form by several governmental initiatives.

Financial investment — 4 billion euros dedicated to AI. Following the report conducted by Fields medalist and then member of Parliament Cédric Villani in 2018, the French government decided to implement a National Strategy for Artificial Intelligence (NSAI), divided into two phases and extending to 2025. More recently, President Macron announced a 500-million-euro budget for the development of AI.

Technological investment. France has developed the “Jean Zay,” supercomputer, combines high-performance computing and AI and now performs at a speed of 37  million billion operations per second.

Attracting talents and developing research and education. The creation of a national research program coordinated by the Inria (the National Institute for Research in Digital Science and Technology), France aims to become an AI research champion.

Ethical AI. Given the close link between AI and personal data, the data protection authority, the National Commission on Informatics and Liberty (CNIL) has also been at the forefront of the reflection on this matter. The CNIL is particularly interested in the issue of generative AI and deep learning, which requires the use of a large amount of data, often collected on the internet, that may include personal data. The authority has launched an ongoing call for contributions, enabling AI actors to give concrete examples of situations they have encountered, processes put in place to comply with current regulations, and, in particular, the way in which data minimization is integrated into the AI creation process.

The European regulatory framework under discussion, may impact French plan for AI. The European Parliament and the Council of the European Union have submitted a proposal to regulate  AI, the so-called AI Act. This proposal would implement risk-based regulation, subjecting AI actors to more or less stringent obligations depending on the risks their operations are deemed to present. A separate would aim to adapt noncontractual civil liability (i.e. tort) rules to artificial intelligence.

Proposals to regulate AI are a source of concern for France and the French companies involved, insofar as some feel that they may stifle innovation by imposing overly burdensome obligations.

Thierry Breton, the French Commissioner for the Internal Market of the European Union, has defended the proposed new regulatory scheme against threats of some CEOs in the technology sector that they would leave the EU market because of the rigidity of the proposal: “No companies will do that, as the EU market is the biggest digital market, and we welcome everyone.”

At the 2023 Vivatech trade show, when asked about the concern regarding the rigidity of the proposed AI Act, President Macron stated: “The priority is to accelerate research, innovation and the creation of a powerful ecosystem, and to participate in regulation, but with the broadest possible scope. The tighter the scope, the more rigidity we create for innovation. The worst-case scenario would be for Europe to invest much less than the USA and China and fail to create leading champions but decide to start with regulation. This is not the scenario I would support.”

What’s next? The trilogue (i.e. informal tripartite meetings on legislative proposals between representatives of the EU Parliament, the EU Council and the EU Commission) period began in June for the EU institutions in view of reaching an agreement by the end of the year. Nevertheless, none of the provisions under discussion will come into force before 2025 or, rather, 2026. This will allow the various actors to look ahead and prepare for their new obligations.

In France, a legislative proposal to regulate AI through copyright protection was put on the table on Sept. 12.

The discussions around the legislative processes both in France and in the EU will be the opportunity for all actors to voice their positions and will have to be followed closely.