OpenAI CEO warns of AI risks and urges strict UN governance

OpenAI CEO Sam Altman briefed the United Nations Security Council today, emphasizing that the development of artificial intelligence must proceed with "extreme caution." Altman pointed out that the scale and performance of deep learning models have continued to break through in recent years, warning that without appropriate governance frameworks, the rapid proliferation of the technology could generate unpredictable societal and security impacts in a short period. The briefing noted that widespread applications ranging from automated decision-making to generative content have already demonstrated AI's potential influence on information manipulation, financial market volatility, and the military sector, calling on the international community to establish risk assessment and regulatory mechanisms concurrently during the research, development, and deployment phases.

Prior to this, multiple countries and international organizations worldwide had already initiated discussions on AI governance. Advanced economies such as the United States, the European Union, and Japan have successively proposed relevant policy drafts focusing on core issues such as data privacy, algorithmic transparency, and accountability. The United Nations also established a special task force on artificial intelligence and machine learning last year, aimed at promoting transnational cooperation and knowledge sharing. Altman's remarks come at the peak of this policy wave, highlighting the need for communication between industry leaders and international decision-makers while reminding nations not to overlook the bottom lines of ethics and safety in the pursuit of technological advantage.

The accelerated development of artificial intelligence has already manifested significant benefits across multiple industries, such as enhanced precision in medical diagnosis, real-time optimization in supply chain management, and improved immediacy in language translation. However, the risks that accompany it should not be underestimated. If generative models are abused, they could produce highly realistic disinformation, further eroding the credibility of public information; in the financial sector, if automated algorithmic trading lacks monitoring, it could trigger severe market volatility; furthermore, some perspectives point out that if future military systems adopt autonomous decision-making AI, it will pose unprecedented challenges to international security. Altman used specific cases in his briefing to urge all parties that while achieving technological breakthroughs, protective mechanisms must be simultaneously established to avoid ignoring potential hazards due to excessive optimism.

Regarding how to implement the principle of "extreme caution," Altman put forward three specific recommendations: first, establish globally unified safety testing standards to ensure that every generation of models undergoes rigorous risk evaluation before public release; second, promote open oversight platforms that allow academia, civil society groups, and government departments to jointly examine the operation and impact of models; third, encourage the allocation of research and development resources toward interpretability and fairness to reduce bias and inappropriate decision-making caused by black-box algorithms. These recommendations echo the non-traditional security threats long monitored by the United Nations Security Council, providing actionable directions for formulating future concrete policies.

In summary, Sam Altman's address at the United Nations Security Council is not only a warning about artificial intelligence technology itself, but also a call to the global governance system. As artificial intelligence gradually permeates all aspects of daily life and national security, how to strike a balance between promoting innovation and preventing risks will become a major issue that the international community must face together in the future. Only within a framework of multilateral cooperation, transparent oversight, and rigorous regulation can the development of artificial intelligence truly serve human well-being rather than becoming a new source of instability. (Source of facts: Central News Agency)