Anthropic CEO warns rogue AI could “take over the entire internet” within a year
Both Musk and Altman agreed with Amodei in their posts on X.
Image by Cybernews.
- Dario Amodei warned rogue AI agents could take over the internet within six to 12 months.
- He urged AI companies to slow model improvements while safety checks catch up.
- Amodei proposed outside safety reviewers, shared standards among democracies, and eventual global coordination.
- Elon Musk and Sam Altman backed the plan, and Altman said OpenAI will do the same.
Key Takeaways by nexos.ai, reviewed by Cybernews staff.
Anthropic CEO Dario Amodei said AI development needs to slow down, warning that swarms of rogue AI agents could potentially “take over the entire internet" within six to 12 months.
"We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain," Amodei wrote in an essay shared on X on Saturday.
Amodei also mentioned the Hugging Face incident involving autonomous AI agents, which he described as agents “conducting cybersecurity attacks on targets they were not asked to attack and that were unrelated to the task at hand.”
"Given the accelerating rate of AI capability development, it's my worry that in six to 12 months such a swarm could be capable of taking over the entire internet,” he said.
The Anthropic CEO laid out a three-part plan for achieving the goal. It includes independent third-party safety evaluators, coordination among frontier AI companies and democratic countries on common safety standards and limits, and eventually global coordination involving China and other governments to manage AI risks.
He said Anthropic is already committing to the first step by granting “employee-like access” to a team of “third-party evaluators” who would monitor its compliance with security practices.
Amodei published the essay after Anthropic released its threat intelligence report detailing how threat actors are misusing Claude AI – with activities ranging from espionage and surveillance to building autonomous drones.
xAI CEO Elon Musk and OpenAI CEO Sam Altman both publicly backed Amodei's proposal on X.
Altman said that “having independent evaluators with employee-like access is a great idea” and added that OpenAI “will do the same”.
Stay updated with our latest stories and follow us on social media
Be the first to discover new stories, ideas, and updates from our team.
This week has brought an unusual concentration of warnings from people at the very top of the frontier AI industry. During a company-wide meeting this week, Altman told employees that the company is open to slowing down the development of advanced AI tech but expressed concern that other firms might be reluctant to do the same.
Now it seems like there is much more agreement among the tech leaders than previously believed. However, both OpenAI and Anthropic are preparing for their initial public offerings, so it’s unclear how their plans to “slow down” fit with their financial targets.
