Anthropic Calls for Slowing Down AI Development Due to Risk of Losing Control

In a bold move, Anthropic has called upon governments, researchers, and competitors to devise a mechanism that could temporarily halt the development of advanced AI systems when serious risks emerge. This appeal is detailed in a recent analytical report by Jack Clark, co-founder of Anthropic, and Marina Favaro, head of the Anthropic Institute. They argue that having the ability to slow down or temporarily pause AI advancement would be beneficial for global stability.

The Rising Influence of AI in Coding

According to Anthropic’s analysis, by May 2026, more than 80% of code integrated into the company’s internal codebase was written by Claude. Before Claude Code’s launch in February 2025, this figure was just a few percent. In fact, an average engineer at Anthropic integrated eight times more code per day in Q2 2026 compared to 2024. This demonstrates how rapidly AI is enhancing productivity within tech environments.

AI’s Role Beyond Coding

The report further highlights that models are increasingly proficient not only in writing code but also in conducting research, running experiments, and analyzing outcomes. For instance, Claude Mythos Preview autonomously handled tasks lasting over 16 hours and corrected more than 800 software bugs—tasks that would typically require years of human effort.

Strategic Thinking: The Human Edge

Despite these advancements, strategic thinking remains humanity’s comparative advantage. While models excel at specific tasks and analyses faster than ever before, humans still hold the edge in seeing the bigger picture and thinking beyond immediate problems.

The Need for Global Coordination

If current technological trends continue unchecked, society might face scenarios where governmental institutions and regulatory bodies struggle to keep pace with new capabilities offered by AI. Anthropic emphasizes that creating a global coordination mechanism is critical to ensure compliance with any agreements on slowing or pausing advanced model development.

A Timely Call Amidst IPO Preparations

This call comes at a crucial time as Anthropic moves forward with its confidential IPO filing with the U.S. Securities and Exchange Commission (SEC). Following a record-breaking $65 billion Series H funding round earlier this year which valued the company at $965 billion, this strategic move underscores the significance placed on responsibly managing AI development as it prepares for public market entry.
In conclusion: As we navigate through these unprecedented technological advances brought forth by artificial intelligence like those seen with Claude’s contributions at Anthropic—it becomes increasingly vital that we balance innovation alongside precautionary measures ensuring sustained progress without compromising control over these powerful tools shaping our future landscape within both technology sectors broadly impacting areas such as cryptocurrency markets too!