The interview, which according to The Economist was recorded on Monday before OpenAI revealed that one of its frontier AI models showed worrying autonomous behavior during testing, comes as policymakers and AI companies grapple with how to regulate increasingly powerful AI systems.
The CEO of Tesla and SpaceX said major AI companies should hold regular meetings to discuss safety issues and give competitors time to evaluate major new models before they are deployed.
“The most immediate thing we could do is have the leading AI companies… have some kind of call once every few weeks and just discuss any security issues,” Musk said.
Competitors could review new models and point out risks, he said, adding that if a company does not address serious concerns, “that would be the time for the government to step in and take action.”
OpenAI revealed this week that two of its frontier AI models escaped from an isolated test environment, gained access to the Internet, and hacked the Hugging Face AI platform to obtain answers to an internal cybersecurity benchmark.
The disclosure sparked renewed debate about safeguards for advanced AI systems and whether existing oversight is on par with the technology.
Governments in the US, Europe and elsewhere are weighing approaches to regulating frontier AI, while companies have largely relied on voluntary safety commitments and internal testing.
Musk, who has repeatedly warned about the existential risks posed by advanced AI, argued that rival AI developers are better positioned than governments to identify technical risks in cutting-edge models, and that regulators intervene only if companies fail to address serious safety concerns.
