Around 400 scientists call for global AI oversight and air-gapping
Technology
Around 400 scientists, including big names like Geoffrey Hinton and Yoshua Bengio, are calling for a worldwide body to keep advanced AI in check.
They say recent security slip-ups show that letting tech companies police themselves is just not enough, and suggest keeping the most powerful AI systems initially air-gapped from outside networks.
Scientists seek strict standards, containment, registry
This push comes a month after a White House deal where companies like OpenAI and Google agreed to voluntary safety steps, but nothing legally binding.
The scientists want strict industry standards, better containment practices, and a public registry of significant security incidents that would eventually be made public.
They are hoping this will help prevent future problems as AI keeps getting smarter.