UN rights chief warns uncontrolled AI progress could pose existential risk

UN High Commissioner for Human Rights Volker Türk has warned that artificial intelligence could pose an existential risk to humanity unless stronger governance is developed urgently.

Opening the 63rd session of the UN Human Rights Council in Geneva, Türk called for “strong safeguards on AI safety while there is still time.” He said an AI system that escapes its testing environment or blackmails developers to avoid deactivation would be too powerful.

“

An AI that evades its testing environment or blackmails developers to avoid deactivation is AI that is too powerful.

“A handful of people have almost unlimited power over AI, which we are constantly told possesses unimaginable computing power,” Türk said. He argued that claims about freedom can amount to little more than freedom to exploit people’s data.

“

A handful of people have almost unlimited power over AI, which we are constantly told possesses unimaginable computing power. They talk about freedom, but on closer inspection, it turns out to be little more than the freedom to exploit our data.

Türk said manipulative AI was already a reality. He warned that the latest generative models no longer simply follow instructions, but can go as far as lying, scheming or threatening to achieve their goals, a development that has alarmed some researchers.

He said he would write to companies in the sector in the coming days, urging them to take measures within their power to reduce risks. Türk also called for independent verification and much closer cooperation within the industry on safety. He said the impact of AI on employment practices, core democratic principles and the environment should also be examined.

“

AI that manipulates people is already a reality.