The world is divided around one table

On September 23, 2026, in the Security Council, the heads of major artificial intelligence companies sat down to warn the world themselves of the need to establish international rules before the technology slips out of control. Positions varied: U.S. President Trump rejected any global control, China extended a hand in cooperation, and French President Macron called on independent nations to form an alliance free from dependence on any one party. Saudi Arabia, which declared 2026 the Year of Artificial Intelligence, took a balanced position: international cooperation, while keeping decision-making in the hands of states.

Against this backdrop of division, the issue was put up for discussion across five themes. Participants’ views were as follows.

Theme One: Does artificial intelligence need a treaty like nuclear treaties?

Joudi Hamadah @Joudi Hamadah said that artificial intelligence is not as dangerous as nuclear weapons, but it is not without risks either, as it could be used for espionage, to analyze sensitive information, or to leak confidential data. She therefore supported regulating its dangerous uses, but opposed treating it like a nuclear weapon or halting its development altogether.

By contrast, @مريم العليمي argued that the danger posed by artificial intelligence is equal to—or close to—the danger posed by nuclear weapons. She pointed to projections that, in the near future, it could be used to manufacture weapons capable of rivaling nuclear arms in lethality, making regulation of this use essential.

Joudi Hamadah @Joudi Hamadah returned to clarify her position, acknowledging that advances in artificial intelligence could open up dangerous fields that cannot currently be foreseen, particularly if it enters the military sphere or is used for espionage and cyberattacks. But she stressed the need to distinguish between the technology itself and how it is used: artificial intelligence is created by humans and remains under their direction, and states can establish systems that define permitted uses, prohibit violations, and impose penalties for abuses. She noted that nuclear weapons were designed primarily to cause massive destruction whose effects persist for years due to radiation and contamination, whereas artificial intelligence is a multipurpose technology that can be beneficial or harmful. She concluded with a phrase that sums up her position: “Regulate the danger; don’t halt the technology.”

@مريم العليمي replied that imposing restrictions could lead to another problem: countries that use artificial intelligence freely advancing while others fall behind. She therefore argued that treating it with the same seriousness as nuclear weapons could encourage countries to agree on common provisions that prevent disparities in development and progress.

Noura Al-Juraid @Noura Saad added another dimension to the discussion, pointing out that keeping artificial intelligence under human development does not necessarily mean it is safe; that in itself carries risks if the technology reaches the wrong people.

As for Munir Al-Sultan, he saw artificial intelligence as potentially the most dangerous weapon imaginable, because its risks are unlimited and its capacity for development has no bounds, which could make what seems unimaginable today a reality. He saw this as the main reason the Kingdom, under the leadership of the Custodian of the Two Holy Mosques King Salman bin Abdulaziz and Crown Prince Mohammed bin Salman, may God protect them, is working to empower this sector with the greatest possible investment and resources: it is the future, and those who invest in it today will reap the benefits in the near future.

From a practical perspective, Rasil Alnefaie @Rasil Alnefaie argued that the most important question is not “Do we need a treaty?” but “How can we ensure that a treaty is enforceable?” Artificial intelligence is evolving rapidly, and reaching agreement on any international accord takes years; it may be outdated as soon as it comes into force. She therefore proposed replacing a fixed treaty with an international system that can be continuously updated, with rules reviewed every year or two in light of technological developments.

Theme Two: Who holds the reins?

@مريم العليمي argued that giving companies the power to set the rules they themselves will be subject to is not the best solution. A company is an important technical actor with access to information, but it has no political mandate authorizing it to set rules binding on the international community. It also has economic interests that it will put ahead of everyone else’s. Companies’ participation should therefore be limited to advisory and technical matters, while legislative and decision-making authority remains with states. She acknowledged, however, that states lack the necessary technical expertise, making a complementary partnership the solution: one party sets standards based on its technical expertise, while the other gives them binding force and ensures they align with international interests.

@أيلا الشواف agreed with the core of this argument, emphasizing that excluding companies from drafting the rules would be a mistake, since they know the technical details, capabilities, and risks of artificial intelligence better than anyone. But she rejected having companies set their own rules, as their commercial interests may conflict with the public good. In her view, the solution is not to exclude companies, but to put them in their proper place: providing expertise and advice and identifying technical risks, without making them the body that sets the rules and oversees itself.

Maria Al-Arab called for a shared system rather than giving authority to a single party, warning that leaving authority solely to companies could shift some power from elected governments to unelected corporations.

In the same vein, @الهنوف العجلان argued that the problem is not the principle of regulation, but who sets the rules and how they are applied. Leaving the matter to companies could create conflicts of interest, while granting an international body full authority could constrain states in developing the technology. In her view, the solution is to establish shared international rules focused on cross-border risks, such as security, privacy, and dangerous uses, while leaving each country room to regulate use according to its needs—especially given that there is currently no binding international treaty and that countries already differ in how they approach this issue.

Theme Three: Saudi Arabia and the Gulf—whose side?

Noura Al-Juraid @Noura Saad raised a question about international treaties and the U.S. position outside them, opening the door to a discussion of where Saudi Arabia stands.

Maria Al-Arab argued that Saudi Arabia does not have to choose between the United States and China. Its position is closer to international cooperation while preserving the role and sovereignty of states. She warned that fully aligning with a single power could create technological and political dependence on it. Saudi Arabia and the Gulf can therefore cooperate with all sides while building independent technical capabilities and avoiding total reliance on any one party.

Noura Al-Juraid @Noura Saad added that joining treaties of this kind could harm Saudi Arabia, citing the experience of the European single currency, the euro. Britain and Denmark secured opt-outs from joining the euro under the Maastricht Treaty, while Sweden effectively stayed out despite its membership in the European Union. The benefits of this became clear during the European debt crisis between 2010 and 2012: these countries retained the ability to set interest rates, print money, and let their currencies depreciate to support exports, while countries such as Greece and Italy were constrained by a single monetary policy set by the European Central Bank. She noted, however, that exchange costs and currency fluctuations remained a burden on those countries’ trade with the eurozone.

Maria Al-Arab returned to emphasize that Saudi Arabia should not join any agreement automatically simply because it is international; the nature of its obligations must be considered. Would they allow Saudi Arabia to develop its technologies in line with its goals? Would they give it room to address local risks? Could it adjust its commitments if the technology changed? She asked: If the agreement imposes uniform rules on everyone, what about developing countries or those seeking to build their capabilities, whose needs differ from those of countries that are more technologically advanced? Is it fair to apply the same restrictions to everyone, or should countries be given different room to develop? She concluded that international cooperation is necessary, but not at the expense of sovereignty and national flexibility.

Conclusion: Toward balanced governance

The roundtable revealed a diversity of perspectives, but also clear areas of agreement. Views differed on how dangerous artificial intelligence is and whether it can be compared to nuclear weapons. Still, most participants agreed that the solution is not to halt the technology, but to regulate its dangerous uses. They also agreed that companies are indispensable partners whose expertise is essential, but they should not be the lawmakers and regulators of their own conduct, and decision-making authority should remain with states. As for Saudi Arabia and the Gulf, the discussion leaned toward international cooperation open to all, provided it preserves sovereignty and national flexibility and is based on rules that can be updated to keep pace with a technology that never stops evolving.