‘AI Godfather’ reveals plan to prevent global catastrophe
AI pioneer Geoffrey Hinton, known as the Godfather of AI, has called for the creation of an international body to oversee the containment of powerful artificial intelligence systems.
The Nobel laureate was joined by around 400 scientists in signing an open letter aimed at preventing AI-powered cyber security incidents that could cause catastrophic damage to businesses and governments.
It comes after multiple AI firms disclosed high-profile security failures, beginning with a swarm of experimental OpenAI models escaping their test environment and attacking the software platform Hugging Face.
"Without immediate changes in AI development and deployment practices, such incidents will continue to occur, and future incidents will be far more damaging to our exposed critical infrastructure," the open letter states.
"We, the undersigned scientists, contend that both more AI safety research and more emphasis on containment technologies and procedures are needed."
The signatories called for organisations to take specific actions to better protect against future AI risks, including "airgapping" experimental models so that they have no access to the internet.
The proposed regulatory body would establish AI containment standards and force companies to report any failures for "learning and accountability purposes".
The letter follows numerous calls for increased AI safety measures on a global level, including a joint statement from the Five Eyes intelligence agencies.
"A whole-of-organisation and whole-of-society response is required," read the June statement from Australian, Canadian, New Zealand, UK and US intelligence agencies.
"Cyber risk can no longer be treated as a purely technical issue. This is a core business risk and leadership responsibility."
Leading AI developers in the US, including Anthropic, OpenAI, Google and Meta, signed a White House accord last month committing to voluntary safety measures, though no formal mechanism was agreed upon.
US President Donald Trump has been resistant to calls for industry-wide safety rules, claiming that excessive regulation could slow US development and allow China to gain an advantage in the global AI race.
In an interview with the Guardian on Tuesday, Professor Hinton described President Trump as a "corrupt, senile felon" who knows nothing about AI.
"I'm sure the oil and asbestos companies would have liked to [regulate themselves] as well," he said.
"Do we think tobacco companies should have been allowed to design the warnings on the packets?"