OpenAI, Anthropic, and Google DeepMind are engaged in ongoing discussions regarding the safety of advanced artificial intelligence models. Chris Lehane, head of global policy at OpenAI, confirmed that these companies are coordinating efforts to mitigate potential catastrophic risks associated with AI development.
The dialogue follows an essay by Anthropic CEO Dario Amodei, who urged the industry to slow down the development of large language models. This initiative received support from Sam Altman of OpenAI, Demis Hassabis of Google, and Elon Musk. Altman also expressed readiness to involve independent auditors for internal safety monitoring.
Participants in the discussions dismissed concerns regarding potential antitrust violations. While Amodei initially suggested seeking government permission for these talks, Lehane stated that such authorization is unnecessary for establishing safety standards.
The companies have previously explored creating an industry-led standardization body. Demis Hassabis had proposed that the US government establish a regulator to oversee powerful models, but the current administration under President Donald Trump rejected these initiatives, viewing them as a potential hindrance to technological leadership.
OpenAI has publicly supported the FRONTIER Act (H.R.9925). This proposed legislation aims to implement federal oversight for the development of advanced AI and mandates that major laboratories allow independent organizations to conduct safety audits of their technologies.
