Next week, OpenAI CEO Sam Altman will appear before the United Nations Security Council in a briefing as governments and heads of technology companies are being asked more questions about the swift evolution of artificial intelligence and how a more powerful system might impact on national security. The meeting will be held during the annual gathering of the United Nations General Assembly in New York, where one of the world’s most prominent AI executives will be directly in front of a body that is tasked with maintaining peace and security internationally.
The Security Council is scheduled to hold an open meeting Wednesday on artificial intelligence and international affairs. The session coincides with growing worries of AI developments, and the fact that many researchers, tech leaders and policymakers are now grappling with how quickly AI is evolving. There have been recent calls to increase coordination between AI companies and governments, especially with the emergence of new systems and their increased abilities being used in more activities.
But an OpenAI spokesperson confirmed that Altman will be in the room. His speech is expected to focus on the need for international cooperation, the role of AI safety and the measures put in place to make sure the benefits of AI are felt globally,” the company said.
OpenAI pointed out to Reuters that “Sam Altman will be briefing on the floor of an open UN Security Council meeting next week.
Altman’s participation coincides with a growing surge of attention on AI safety issues. In recent years, some key technology leaders have voiced a desire for increased protections, more coordination, and more stringent parameters with the rollout of increasingly strong AI systems. They have expressed fears that the more sophisticated systems might become harder to monitor and control as they become more powerful.
It’s no longer just a tech vs. lab debate. The governments have started to consider AI in the context of national security, economic competitiveness, cyber operations and military applications. This has complicated international coordination, as countries have varying priorities for technological development, for regulation and security.
France leads the council as its President for the month of September, and is calling the upcoming Security Council session. The meeting will be headed by French Minister of European and Foreign Affairs, Jean-Noël Barrot. The debate has been surrounded by concerns about the development of safe and responsible AI and its potential for malicious use.
The meeting also highlights the growing way AI is a part of the international peace and security conversation. AI systems are now employed in various fields, including information management, cyber security, scientific research, and business transactions. Meanwhile, governments and experts are looking at the impact more autonomous systems might have on critical infrastructure, military decision-making and misinformation.
Senior officials from Anthropic also could be in attendance at the meeting, diplomats have said, though at the time of the report, no indication had been made about the company’s attendance. With multiple leading AI firms in attendance, the conversation might yield a variety of insights from a sector that’s making significant strides in technology while also facing increased concerns about safety.
The Security Council has already discussed the global aspects of artificial intelligence. The body held its first meeting on the risks of AI in 2023. At the time, China said it did not want the technology to develop without proper management and the United States raised questions about how it might be used for censorship and re-pression.
This shift in focus aligns with the ongoing rapid progression of AI capabilities. Advanced text, images and software can be produced, along with advanced research and increasingly complex tasks, using modern systems. As their utility increases, they are becoming commonplace, but the same properties have also sparked questions of security, accountability, and what are the bounds of autonomous systems?
A major challenge for policy makers is how to establish safety standards that can be agreed on across borders. AI companies have global reach, and regulations for AI are typically developed by specific governments or regional organizations. The variation in national rules can therefore pose problems when policymakers try to establish common rules for high-capable AI systems.
International cooperation may go some way to solving some of these problems, but it is not easy to find consensus among governments with contrasting economic and strategic interests. Countries are negotiating protection measures to mitigate the risks of technology at the same time as they are competing for technological leadership. These more geopolitical issues can be addressed alongside technical questions of AI safety in the Security Council.
Altman has been talking about the dangers of super-intelligent AI for over 10 years. He stated in an essay, “Machine Intelligence, part 1” (2015) that the “development of superhuman machine intelligence is probably the greatest threat to the continued existence of humanity”.
The earlier warning is now resurfacing as AI systems have transitioned from experimental research laboratories to the consumer market and professional applications. Things have moved on since Altman made those comments, and the technology sector now has generative AI as a key component of consumer and business technology. As AI continues to evolve, businesses are facing greater scrutiny to prove their safety measures are advancing alongside technological advancements.
Altman’s presence at the Security Council is a chance for OpenAI to share its strategy on how to ensure AI remains safe and collaborates internationally directly with government representatives. His comments will also be closely scrutinized as OpenAI is one of the companies working on some of the world’s most advanced AI systems.
It’s an important and delicate topic for the industry. Businesses are pushed to create more powerful models, and to overcome worries of misuse, reliability, and potential side effects. The question of the amount of self-regulation or government regulation and international agreement necessary to regulate technology is still under debate.
