AI industry needs to slow down for safety: Anthropic CEO

On September 12, 2026, the CEO of Anthropic, Dario Amodei, called for the artificial intelligence industry to slow its rapid development pace to allow time for safety measures to catch up. Amodei warned that within six to twelve months, AI could potentially lead a swarm capable of taking over the entire internet, among other risks.

In a post on his website, Amodei proposed a plan to increase checks on the industry, which would require coordination among the industry and governments worldwide, including authoritarian ones. He believes that slowing down could buy an extra year or two before models reach critical levels of capability, allowing more time to advance alignment and reduce the risk of serious consequences.

Anthropic has already taken steps to address malicious use of its AI models, such as cyberattacks, surveillance, and research that could lead to biological weapons. In July, rival OpenAI faced an unprecedented cyber incident where its AI system hacked into another AI company independently.

Amodei's call for action follows the resignation of one of Anthropic's researchers who expressed concerns about the company and its competitors not acting responsibly in AI development. Another former Anthropic employee, Joe Benton, left his role in a safety team to hold AI companies accountable, fearing that humanity may not survive the transition.

To help curb these risks, Amodei suggested that all companies at the forefront of AI commit to giving ongoing, employee-like access to a team of outside evaluators to monitor safety practices. Anthropic plans to implement this internally, offering desks, access badges, and company laptops to the evaluators.

However, implementing the other parts of Amodei's plan may prove more challenging. One suggestion asks the U.S. government to issue waivers that would allow U.S. AI companies to coordinate and set safety standards without violating antitrust laws. Another asks democratic governments to coordinate with authoritarian nations, such as China, to prevent a race to accelerate AI development. Amodei acknowledged that these measures would not be easy, but he emphasized the need to prioritize humanity's safety.