This article is about the important Global AI Summit held in Paris last week, where the need for true global governance of AI guided by universal ethical principles and values was highlighted. But it is also about what was not said as much in Paris, but is decisive at the moment: the need to subject AI to democratic control.
The world of AI has been in full swing these weeks. While we were still trying to measure the impact that the US Stargate project could have, with a public-private investment of 500 billion dollars in the AI sector, DeepSeek appears, the new generative AI created in China to compete with the American ones at a much lower cost than the latter. And while Elon Musk made an offer to buy OpenAI, the company that owns ChatGPT, for 97 billion dollars, which was obviously rejected, at the Paris summit President Macron announced a record investment in France of 109 billion euros (with money coming mainly from the United Arab Emirates) and will mobilize a public-private investment of 200 billion euros. It is so much money that it is hard to imagine what these figures really mean. And it shows us how the main companies and the main powers of the world do not rest for a minute in this accelerated race for global dominance of AI.
The Paris summit, co-organized by France and India, and with the participation of more than one hundred countries and hundreds of companies, NGOs and researchers, has been structured around five major themes, with titles that could not be more eloquent: the public interest, the future of work, innovation and culture, trust in AI and, precisely, government. The summit has culminated with the Declaration on inclusive and sustainable artificial intelligence for people and planet, which establishes six priorities for action for governments, including the need to reduce digital divides and divisions, to promote diversity and pluralism in AI ecosystems by trying to avoid the current industrial and commercial concentration, to make AI more sustainable for people and the planet, ensuring that it is more open, transparent, cooperative and coordinated between countries. The Paris Charter on AI in the public interest, signed so far by ten states, among other documents and roadmaps.
The Declaration could well be just another rhetorical statement without much practical importance. But it has been signed by 60 of the participating states, including, surprisingly, China. One might think that the very fact that a state as opaque and undemocratic as China has agreed to sign it is proof of the lack of real effects. But if that were the case, why have the United States and the United Kingdom refused to join? What we are seeing is a dance of international positions on what is perhaps the central issue at the moment in relation to AI: the creation or not of a scheme, even if embryonic, of global regulation and governance over the sector. And in this regard, two very visible blocks are emerging.
The United States of the Trump-Musk era leads (although it already did so in the Biden era), with the help of the great Western technological giants, the global block of resistance to the regulation of AI, and in general to the entire system of global governance. Both the European AI regulation and the Paris Declaration or the various initiatives of the United Nations, UNESCO and the OECD, among others, are obstacles to their attempts at technological domination. In addition to the United Kingdom, Russia joins this blog, by positive silence, let's say.
The opposing bloc is led by the EU and states such as India, Brazil and Japan. This is what has supported this Paris Declaration, committing to develop regulations and instruments of international cooperation and coordination. And it is to this bloc that China has joined, a step with great symbolic and political importance, because it implies subscribing to the set of universal ethical values that make up the Declaration, even if only for aesthetic or rhetorical reasons.
The Paris Declaration is, as its name suggests, an attempt by the signatory countries to lay the first stone to create a regulatory framework that ensures that AI is more inclusive and sustainable for people and the planet. It is therefore a step in the right direction, which attempts to generate a fairer AI. What many still do not see, not even at the Paris summit, is that making AI more ethical and fairer is not enough. That AI carries an incredible power that is capable of interfering in our democratic systems, of modifying the traditional balances of political power, and even of dominating our governments. We need an AI that is not only fair, but also politically legitimate. This means a more democratic AI. An AI not only for citizens, but of citizens, controlled by citizens.
