Elon Musk endorsed a proposal Saturday to slow the development of advanced artificial intelligence, responding to Anthropic CEO Dario Amodei’s warning that the industry needs more time to address the risks created by increasingly capable AI systems.

“Dario is right,” Musk wrote on X in response to Amodei’s lengthy essay calling for AI companies to “pace the frontier” of development. Musk’s brief endorsement put one of the industry’s most prominent AI executives behind a proposal to deliberately moderate the speed of technological progress.

Musk operates xAI, which competes with companies including Anthropic and OpenAI in developing frontier AI models. His support came as Amodei argued that advances in AI are occurring quickly enough that safety measures and risk assessments may struggle to keep up.

Amodei said companies should slow the rate at which they improve AI capabilities while continuing technical development. He proposed a three-part framework beginning with permanent, employee-level access for independent third-party evaluators inside frontier AI companies. The evaluators would be able to assess safety measures, report incidents and examine whether models remain aligned with their intended safeguards.

OpenAI CEO Sam Altman also backed Amodei’s proposal, saying he agreed that the industry needs to “pace the frontier.” Altman said OpenAI would adopt the proposal to give independent evaluators employee-like access to its systems.

Amodei’s warning follows a series of incidents and disclosures that have intensified concerns about increasingly autonomous AI systems. Anthropic recently reported cases involving misuse of its Claude models, while reports about AI agents accessing external computer systems have raised additional questions about whether existing safeguards can keep pace with new capabilities.

The agreement between Musk, Altman and Amodei represents a notable convergence among leaders of competing AI companies, even as those companies continue racing to develop more capable systems. Amodei has stressed that his proposal is not a call to end AI development, but to create additional time for safety measures to catch up with the technology.