At UN, tech chiefs urge caution in AI development

The United Nations Security Council convened an emergency meeting on Wednesday to discuss the rapid development of artificial intelligence (AI) and the need for caution in its advancement. Representatives from leading AI firms, including OpenAI and Anthropic, addressed the council, emphasizing the importance of responsible AI development.

Sam Altman, CEO of OpenAI, warned that AI systems are becoming more capable and autonomous, potentially moving faster than institutions and concentrating power in the hands of a few. He urged governments to collaborate on shaping the future of AI technology and cautioned against embracing excessive technological risk for the sake of significant benefits. Altman also highlighted the potential for AI to pose a risk to humanity if not managed properly.

Anthropic's CEO, Dario Amodei, echoed Altman's concerns, stating that as AI models become more powerful, more stringent safety standards are necessary. He pledged that Anthropic would slow down the release of AI technologies until they are deemed safe. Other industry leaders, such as Hugging Face's Clement Delangue, acknowledged the dual nature of AI, acknowledging its potential for misuse while also highlighting its role in defending against cyberattacks.

The meeting comes as the United States prepares to host Chinese President Xi Jinping, with AI set to be a key topic of discussion. OpenAI CEO Sam Altman advocated for a "middle path" between blind optimism and extreme caution, urging against both overconfidence and undue pessimism. Other prominent AI researchers, including Canadian computer scientist Yoshua Bengio, called for productive action in response to growing concerns about AI's rapid advancements.