“AI” a crisis in global governance (By Seydou KA) October 6, 2026
“Is the effectiveness of technical action necessarily a source of moral progress? “. This subject proposed for the philosophy test during the 2025 Senegalese baccalaureate refers to an often quoted phrase from Condorcet dating from 1787 according to which “the more civilization spreads on earth, the more we will see wars and conquests disappear, like slavery and poverty”.
Indeed, one of the great dreams of the Enlightenment was to believe that the dissemination of modern knowledge and scientific progress should make the human species better. We know that this promise was never kept. Our era is experiencing dazzling technological development; However, wars and conquests have not stopped, poverty has not diminished and even slavery has only been banned in regulations, not in practice. Thus, the explosion of the atomic bomb above Hiroshima is both an index of scientific and technical progress, and proof of a moral regression of civilization. Worse, as recent history has shown, the greatest acts of barbarity have not been committed by particularly uneducated beings.
Contrary to the illusions maintained by certain supporters of the Enlightenment, Rousseau, from his first writing, the “Discourse on the Sciences and the Arts”, separated himself from his philosopher and Encyclopedist friends, by renouncing the belief that the diffusion of works and techniques will make humanity better. Far from contributing to the advancement of morality and increased benevolence for others, he argues, the improvement of the sciences and arts harms moral progress. “We can be men without being learned,” concludes Rousseau in “Émile, ou de l’education”.
These examples provide information on the necessary separation between research itself and its consequences, positive or negative. The secret of the fission of the atom was discovered within Western culture, but the decision to drop the atomic bomb on a few hundred thousand Japanese was also made possible there, probably thanks to the same mechanism of fragmentation and dissociation between ends and means, between morality and knowledge.
Today, humanity once again faces a turning point. The dazzling progress in generative AI raises as many promises as concerns. This summer, the big bosses of American Tech sounded the alarm and called for better control of this technology. The risks of out-of-control AI are real. Some even fear “an extinction of humanity”. Most worryingly, the world’s top leaders do not appear to have taken these warnings seriously. Apart from the salutary position of Pope Leo XIV and UN Secretary General António Guterres, few political leaders seem to be concerned about the potential consequences of AI without human control. As an exception, we can cite the German-Canadian initiative “for AI in the service of humans”. For the rest, geopolitical rivalries have taken over regulation. Donald Trump rejected the warnings of the big bosses of American Tech, thinking that any regulation would leave the field open to the Chinese rival. However, the fundamental challenge today is to achieve global governance of AI. The previous generation had the intelligence to agree to put in place mechanisms for mutual control of nuclear weapons and the Non-Proliferation Treaty (NPT). Even on this register, we are today witnessing a dismantling of the treaties of the Cold War era which, despite their limits, had the merit of avoiding a devastating nuclear war for humanity. Ultimately, AI is not the problem. It is the crisis of global governance that is worrying. And history has shown us that when geopolitical rivalries prevail over cooperation between great powers, it can only lead to disaster.
