Anthropic, the company that created the Claude artificial intelligence model, publicly defended the possibility of slowing down or even pausing AI development, a decision that has generated debate in the technology sector. This position comes in a context where the company already uses its own AI systems autonomously to build the next generations of technology.
In May 2026, more than 80% of the code integrated into Anthropic's production base was written by Claude autonomously, with engineer productivity skyrocketing as a result of this automation. The article warns that we are entering a phase where AI tools are beginning to participate increasingly deeply in building even more powerful systems, something that raises questions about the human capacity to supervise, test, and control technological development.
Anthropic recognizes that an isolated company cannot simply stop development, as that would mean handing advantage to competitors. Therefore, the company advocates for the need for international coordination to establish collective limits. The article questions whether any major power would be willing to be the first to slow down, in a scenario of global technological race.
The author does not advocate stopping progress, recognizing that AI could produce extraordinary advances in science, medicine, and education. However, it warns that progress without prudence may be merely dangerous acceleration, and that reaching too quickly a place that is not yet known how to govern may mean there is no "pause button" effective enough.
