Analyses & Studies
AI action plan outlines how to place France 'at the cutting edge'
This technology represents a potential source of growth and jobs, according to the report submitted to Emmanuel Macron on March 13.
What should France do about artificial intelligence? In response to questions about this technology, the AI Commission submitted a report to Emmanuel Macron on Wednesday, March 13. The ton is "optimistic and positive, even if we know there are risks," said Anne Bouverot, the president of the Ecole Normale Supérieure's (an elite higher education institute) board of directors. Together with economist Philippe Aghion, Bouverot co-chaired the work of some 15 experts commissioned by the prime minister's office in September 2023.
Back in the spotlight since the launch of ChatGPT at the end of 2022, AI is presented as an "inescapable technological revolution," comparable to electricity or the automobile. If France were to overlook it, "we could not only miss out on the AI economy, leading to the increasing capture by others of our economic value, but also see the weakening of other business sectors," the authors warn, pointing to a risk of "economic downgrading." If nothing is done, we risk seeing the ship sail without us, said Bouverot, referring to past waves of digital innovation dominated by American companies.
Faced with these crucial challenges, however, the commission asserts that France can be "at the cutting edge" of AI. "It's an opportunity for France, for the economy and for working people, provided we prepare for it," assured the co-chair. In an "ambitious action plan," she proposes investing €5 billion a year for five years, in particular to finance the sector and the deployment of the technology in companies, but also to train employees and more.
While the report endorses the values of "sovereignty," "social dialogue" and "responsibility," it also aims to remove certain "obstacles." And a number of recommendations could prove controversial: facilitating the establishment of data centers in France; easing certain authorization procedures for the use of personal data, notably in health care or by the police; the creation of a €10 billion "AI fund"; the notion of reforming taxation to finance innovation; and increasing the use of AI in public services, including education.
"Artificial intelligence is a potential source of growth and jobs," according to Mr. Aghion. After years of "secular stagnation" following the "thirty glorious years," "AI could help us regain high growth rates through two effects: by increasing our productivity and by increasing our ability to generate new ideas," explains the report.
The productivity gains associated with AI over ten years could lead to "an increase in GDP ranging from 250 billion to 420 billion euros by 2034," the report estimates. However, this effect, "temporary," would diminish as AI becomes adopted by everyone.
On the employment side, they estimate that "the effects of AI will generally be favorable." "The displacement effect," whereby automation destroys jobs, would be less strong than "the productivity effect," which improves the quality-price ratio of products, creates demand, and leads to hiring.
In the short term, the report also proposes the creation of a France & AI investment fund, endowed with 10 billion euros, of which 7 billion euros come from large French companies and 3 billion euros from public funds. Its goal: to invest in AI startups, either at their inception or as they grow.
On the regulatory front, the commission aims to establish in France a "public and private evaluation ecosystem for AI systems," involving regulatory authorities from various sectors. The objective is to specify standards for "reliable AI in the medical field," "unbiased AI in recruitment," etc. A "AI council" would be dedicated to advising the government and composed of around "fifteen qualified individuals," resembling the members of the commission.
Furthermore, at the global level, there is advocacy for "global governance of AI": a "global AI organization" to assess and regulate systems; an "international AI fund for the common good"; and a solidarity mechanism, "1% AI," to share computing power with developing countries. The text appropriately suggests that this "World AI Organization" could be established in Paris and created during the international summit on the subject that France will host, either at the end of 2024 or the beginning of 2025. However, Paris will need to negotiate with nations that also aim to showcase themselves at the forefront of this technology and were present at the London summit in November.
