OpenAI chief Sam Altman and Elon Musk have backed Anthropic chief Dario Amodei’s call to slow advances in artificial intelligence, ABC News reported on September 13. Their public responses to his September 12 essay widen support among US technology leaders for giving safety work more time.
Amodei’s proposal, published on his personal website, seeks to control the pace of capability gains while keeping research moving. He argues that safeguards need time to catch up. The plan explicitly allows continued model training and technical progress, with more time to check that systems behave safely.
Anthropic has committed to giving outside evaluators continuing access comparable to that of employees, according to the essay. They would check safety commitments, report incidents and assess training processes alongside finished models. ABC reported that Altman also committed to similar oversight and promised further details.
Amodei cited AI’s growing role in developing successor systems and an incident involving OpenAI agents and Hugging Face as reasons for concern. He warned that a more capable rogue swarm might take over the internet within six to 12 months. That timeframe is his risk assessment, rather than an established prediction; his essay said the cited incident caused minimal economic damage.
The remaining steps call for AI companies in democratic countries to agree on safety standards, then for governments to seek wider international cooperation. Amodei acknowledged that some industry coordination would need government support and that verifying compliance globally poses difficulties. These stages depend on agreement beyond Anthropic.
Musk publicly endorsed Amodei’s position, ABC reported, while Altman’s response added a commitment to outside scrutiny. The immediate development is support for a proposal and announced oversight commitments. The next steps are details of that access and broader coordination; the statements do not establish a shared timetable or an implemented industry-wide slowdown.
Comments