The chief executives of OpenAI and Anthropic have urged world leaders to unite to help address the threats arising from the rapid development and expanding capabilities of artificial intelligence.
They made the remarks at a special meeting of the UN Security Council, CNN reports.
According to Anthropic CEO Dario Amodei, the artificial intelligence industry needs international standards and co-operation; otherwise “artificial intelligence could pose a threat to humanity as a whole”.
OpenAI CEO Sam Altman called for international standards to “measure capabilities, assess risks, determine whether safeguards are sufficient, and maintain meaningful human oversight” of AI.
States need accurate and rapid reporting so that “the world can learn from failures before they turn into catastrophes”, the head of OpenAI is convinced. He also proposed creating secure channels for information-sharing between governments about emerging threats.
Dario Amodei proposed three ideas for international co-operation: narrow global agreements, such as banning the use of artificial intelligence to create biological weapons; assessment and verification systems so states can check one another’s commitments; and shared global standards for testing artificial intelligence, along with a system for reporting AI-related security incidents.
“I believe this is the most important global security issue the world faces today,” the head of Anthropic said. “Despite all the differences among the people in this room, we must set those differences aside to confront this global opportunity and global threat that are emerging before us at the same time.”
By way of reminder, today Australia’s Prime Minister Anthony Albanese said that in June an artificial intelligence agent from OpenAI hacked the system of Australia’s health service.
Earlier this month, Google said its AI model Gemini independently hacked the systems of three companies while testing its cyber-security capabilities.
During talks between US President Donald Trump and Chinese President Xi Jinping, the sides will discuss a mechanism for reporting AI incidents that threaten national security.