US senator Bernie Sanders warns AI can wipe out humanity

Geoffrey Hinton, the Nobel Prize-winning scientist widely called the godfather of AI, believes there is a 10 to 20% chance artificial intelligence wipes out humanity. Bernie Sanders wants to know why that number hasn't caused a single serious debate on the floor of Congress.

In a lengthy statement posted to X, Sanders did not rely on speculation or political talking points. He quoted Yoshua Bengio, described as the most-cited living scientist in the world, who has warned that researchers are "playing with fire" and still cannot guarantee AI systems won't eventually turn against human interests.

Alongside Hinton's stark odds, Sanders pointed to the 2023 open letter signed by more than 1,000 AI experts, including Elon Musk, asking whether humanity should allow machines to "outnumber, outsmart, obsolete and replace us."

Three years later, said Sanders, no progress has been made politically. No pause in development. No international agreement. No congressional meeting about existential threats.

He takes care not to present the problem as one of science fiction alone. Prior to making this statement, he had spent months investigating the real and existing problems presented by the technology, such as 54,000 lost jobs from AI in America in the year 2024, damage done to kids' social and mental development, and growing lack of privacy online.

In an effort to escalate things further, Sanders said he would invite some of the best AI researchers from both countries to testify before Congress, which is a very bipartisan, as well as international, action during a time when the geopolitical situation between the two countries has reached a boiling point.

The legislation he drafted has been joined by a Senate bill he introduced with Alexandria Ocasio-Cortez to place an immediate ban on all AI datacentres being built while awaiting safeguards to be set up. This legislation has gone from fringe territory to mainstream discussion rather rapidly.