US Senator Bernie Sanders has called for US President Donald Trump and Chinese President Xi Jinping to start talking about a treaty to freeze the development of superintelligent AI and prohibit it. His comments followed those of the CEOs of Anthropic, OpenAI and xAI who urged a slower pace of AI system development.

The calls by Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and xAI CEO Elon Musk to slow down the development of AI were a ‘start’, but stronger action is needed, Sanders said. He cautioned the growing strength of AI systems could become dangerous if they can run without humans.

Sanders calls for AI treaty

In a post on X, Sanders stated that both Amodei, Musk and Altman now realise that the industry must “slow down and pace the frontier of AI”.

“That’s a start, but it’s not enough,” Sanders said. “When you are racing towards a cliff, you don’t just ease up on the gas pedal. You hit the brakes,” he added.

Sanders called on Trump and Xi to address at the upcoming AI summit an international agreement. The treaty should temporarily halt the development of advanced AI and prevent the development of superintelligent systems, he said. In the past, the senator has expressed worries about AI gaining capabilities that surpass those of humans and becoming autonomous.

What AI CEOs said

The current debate was triggered by a blog post by Dario Amodei, who voiced his concerns about AI. Anthropic’s CEO raised concerns over the following threats: control of AI systems, cyber attacks, bioterrorism, and significant economic disruption.

Amodei said that if companies focus on improving their capabilities at a faster pace than safety, there could be more risks as they compete with each other.

“We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote, while adding that AI progress would still remain rapid.

Sam Altman agreed with the need for greater caution. The OpenAI CEO stated that the company had been discussing the rate at which AI is advancing and supported the concept of independent evaluators having full access to AI systems.

Elon Musk also agreed with Amodei. The xAI chief just replied to the Anthropic CEO’s post, stating, “Dario is right.”

Sanders wants permanent ban

Apart from that, Sanders has submitted a bill to permanently ban superintelligent AI. The idea is to limit the creation of systems able to function without human oversight.

The debate follows the tide of AI firms pouring billions into developing more powerful models. Sanders believes the governments should take action before these systems get too big to manage.

Source link