Intervention by the President of the Republic via videoconference on the occasion of the Seoul Summit on Artificial Intelligence.

France Emmanuel Macron · President of France

21 May 2024 · Seoul Summit on Artificial Intelligence

Context
Macron addressed the Seoul AI Summit via videoconference on May 21, 2024, outlining France's priorities for international AI governance ahead of the Paris AI Action Summit in February 2025.
Main themes
International AI governance, artificial intelligence regulation, AI innovation and investment, open AI architecture, citizen protection from AI risks, militarization of AI, intellectual property protection, environmental impact of AI, disinformation and cybersecurity, global AI summits, sustainable development and AI, fragmentation of AI governance
Duration
~11 min
37th percentile
Words
1,260
37th percentile
Sentences
33
20th percentile
Readability
27
12th percentile
Grade
18.7
95th percentile
Syntax
3.49
94th percentile
Vocabulary
2.06
19th percentile

Translated from French

Content
Actors
Style

The broad sphere of national life a sentence belongs to. Every sentence has an Arena, so it is the denominator the other indices are read against.

Other
  • Geopolitical 36%36%
  • Technological 30%30%
  • Ceremonial 18%18%
  • Economic 15%15%
  • Social 0%0%
  • Environmental 0%0%
  • Institutional 0%0%
  • Not in focus 0%

33 sentences in 14 paragraphs

Hello, thank you very much to everyone.

First, I would like to thank President Yoon, thank you very much dear President, and Rishi Sunak, Mr. Prime Minister, for organizing this meeting and thank all the colleagues, heads of state and government, and business leaders and experts who have taken the time today for this essential exchange. It indeed extends the momentum of the Bletchley Park Summit last November and allows us to revive the discussions held during the founding of the Global Partnership on Artificial Intelligence initiated by Canada in November 2018.

You have all stated it very clearly, and it is widely reflected in the interventions: artificial intelligence is not just an economic and technological revolution – it carries the potential for a profound paradigm shift in our societies, that is, our relationship to knowledge, our relationship to work, our relationship to information and thus to democracy, our relationship to culture and even to language. And quickly, we will have to adopt it within our companies if we do not want to lose the battle for productivity, but also integrate the good and bad uses within our democracies.

For all these reasons, it has clearly become a political issue, and it can be for the worst or the best of reasons, if I may say so, and that is why having an integrated international conversation that allows for the inclusion of technological leaders, users, but also the governors of the planet, seems essential to me. That is why we have accepted the responsibility to extend the momentum initiated by the United Kingdom and Korea by hosting in France, on February 10 and 11, 2025, a summit for action on AI, for which I count on your commitment.

We will get to work in the coming days, and my special envoy, Anne Bouverot, will gather our main partners on May 28 in this regard, to build, fundamentally, an agenda around our dual responsibility towards AI, and this corresponds to what has just been said - that is, both to share, innovate, and continue to accelerate, and to protect against misuse and also integrate the direct and indirect consequences of AI.

First, therefore, to share, to open up. It is about how to ensure that the massive adoption of artificial intelligence can fulfill the initial promise of digital technology, which is individual and collective emancipation, and thus progress, knowledge, and behind that, economic growth. To achieve this, we need an investment agenda to promote innovation and avoid concentration - in doing so, we strengthen the freedoms to undertake, create, and learn. We also need to encourage innovation even more strongly in favor of the common good and the sustainable development goals, whether it concerns the environment or health, by also building a fully "open" architecture of artificial intelligence; I know how much some of you – and Yoon just referred to it – care about this, and to work towards a technology accessible to all.

I do not want to be longer here, but for me, the first part of our action agenda for AI is indeed how to invest more, how we can accelerate innovation, dissemination, and how to ensure that there is a foundation of "open" architecture to allow for the possible development of our main objectives.

Secondly, our role is clearly to protect citizens from the risks generated by the very rapid development of AI, that the riskiest uses are fully prevented, which was much discussed in the UK last year; this is the guiding thread of the European regulation on AI adopted last March. And very clearly, impossible uses of AI are emerging, the militarization of AI and others; I believe we need to integrate this from the very beginning. We need to have a global conversation on this subject from the outset if we want to be effective and thus say that we have a form of international public order that must prevent certain basic misuses. And then, there are things that we must also be able to integrate very quickly and calibrate: the protection of intellectual property, the preservation of everyone's data, the environmental impact of AIs, the conditions of remuneration, the proliferation of cyberattacks and disinformation campaigns. So, we can see that protection will also require a somewhat different approach from a classic approach, that is, to have a stress test approach, by scenario, and to agree on a common agenda.

For this, and I will conclude with this, we need a new international governance of artificial intelligence, and I believe that the great risk would be to have regional governances and that, fundamentally, there would be a fragmentation of the conversation and preferences because they would create a form of disparity and hinder our ability to invest and innovate together and create different collective preferences, while if we want to be effective, we need to build a true common international foundation. I believe that this new international governance, of which we have laid the foundations with this Global Partnership on Artificial Intelligence, is based around the triptych of science, solutions, and standards.

Science first, we must constantly seek to build a robust scientific consensus together. On this, I want to pay tribute to the work entrusted to the expert group led by Yoshua BENGIO, and whose final results will arrive next February, but whose provisional results have already been released and are already referenced, allowing for a panoramic view of the risks and opportunities. I believe it is important that this is scientifically based each time, founded, that there may be discussions and disagreements, but that it should not be the subject of a politically biased debate.

The second element for global governance is to unite our forces to create technical solutions open to all, thus allowing for the auditing of models, determining the right risk assessment standards, and evaluating their environmental impact. I think it is essential that our different institutes for evaluating AI models can share their results, especially with countries that are equally concerned by the issues but are not currently equipped with the same analytical capabilities. Companies must also be able to contribute, but I think it is very important to have these technical solutions open to all; otherwise, there will not be a truly common governance.

And finally, we need to build international standards for AI to have common rules that are fully compatible and do not create distortions between spaces, or additional costs for businesses. It is also for this reason that I believe strongly in our ability to include all geographical areas and all the major technological and economic powers of the planet. This governance should not be done by excluding one or another; we must really avoid a fragmentation of regulatory frameworks that would benefit no one.

So, these three objectives for me – science, solutions, standards – are at the heart of the revival project of the Global Partnership on Artificial Intelligence, of which we are all a part, alongside major partners like India, Brazil, Senegal, about thirty countries in total, soon to be joined by others. A new stage is opening, but it is around these principles that we will build and thus carry out this work in view of next February.

I believe I have exhausted the time allotted to me, and I do not want to take longer, but the objective was really to thank the colleagues and business leaders, contributors who spoke today and to try to outline the main objectives for the summit next February. Thank you very much!