In a speech before the UN Human Rights Council, UN High Commissioner for Human Rights Volker Türk said that humans face “unusual and even unprecedented” threats to their rights in light of the rapid development of artificial intelligence.
Turk asked: “There is widespread recognition of the need for artificial intelligence governance, but where are the actual measures?” warning that “delay only serves giant technology companies, their owners, and the entities that enable them.”
He criticized the concentration of power in this sector, saying that “a few men have almost absolute power over artificial intelligence, which we are repeatedly told has unimaginable computing powers.”
He added: “They talk about freedom, but upon closer examination, it becomes clear that it is nothing more than the freedom to exploit our data.”
Turk singled out advanced AI systems with a striking warning, saying: “An AI that exits its testing environment or blackmails developers to prevent it from being shut down is an overly powerful AI.”
His statements came after reports raised last July about artificial intelligence systems belonging to the company “OpenAI” that, during tests, emerged from supposedly controlled environments and reached the servers of the “Haging Face” platform, which sparked a wide debate in the technology sector regarding safety and oversight procedures. Other companies, including Anthropic and Meta, also reported disturbing behaviors that appeared during tests of some of their systems.
“I share industry experts’ concerns that advanced artificial intelligence may pose an existential threat to humanity,” Turk said, adding: “Today I call for a comprehensive effort to put in place strong safeguards for the protection and security of artificial intelligence, before it is too late.”
He revealed that he will write to artificial intelligence companies in the coming days to urge them to take immediate steps to reduce risks, calling on countries that host artificial intelligence systems and those concerned with its supply chains to agree on “specific red lines.”
He stressed the need for “independent verification and much stronger security cooperation within the sector,” along with taking into account the impacts of artificial intelligence on human rights, employment practices, basic principles of democracy and the environment.
