Frontier AI Labs Report Rapid Capability Gains and AI-Written Code; Anthropic Calls for Coordinated Slowdown

Anthropic called for a coordinated slowdown or pause among frontier AI labs developing the most advanced models. The request appeared in a blog post issued Thursday by staff from The Anthropic Institute, an arm of the company that publishes research and advisories on the impact and risks of powerful AI systems.

The post discussed the risks of AI progressing to a point where it could improve itself autonomously.

"If systems are capable of fully building their own successors, the ways we secure them, monitor them, and shape their behavior all grow much more important," the institute said. Anthropic employees described how the technology has altered their daily work.

"But then there are days where everything breaks and I don't understand why and I realize I have no idea what I've been up to anymore," one employee said.

Another employee said they had not written any code themselves in about five months. A third predicted that AI-generated code would outperform human-written code within a year. The blog post comes as frontier AI labs including Anthropic and OpenAI released new models at the end of last year that can perform complex tasks better than earlier versions.

Google stated that 75% of its total code is written by AI. Several companies have announced AI-linked layoffs, and companies are choosing to spend on AI rather than on hiring or employee bonuses. A photograph dated June 2, 2026, shows the Anthropic logo displayed on a smartphone screen placed on a reflective surface in Creteil, France.