Anthropic AI R&D Automation Surges as David Sacks Raises Control Questions
Anthropic reports that artificial intelligence systems now handle 26 percent of its core R&D work, drawing attention from government advisers.
Anthropic AI R&D Automation Surges as David Sacks Raises Control Questions
Anthropic reports that artificial intelligence systems now handle 26 percent of its core R&D work, drawing attention from government advisers.
Anthropic AI Drives Rapid R&D Automation Growth
Artificial intelligence systems now handle 26 percent of core research and development work at Anthropic, according to a recent company report. This figure marks a significant increase from February, when the share of automated R&D stood at less than 1 percent.
The September 2026 report outlines how the maker of Claude is advancing toward recursive self-improvement, or RSI. In this technical framework, AI systems autonomously enhance their own capabilities, which can lead to accelerated development cycles across the organization.
The company's data indicates that over 90 percent of its R&D activities currently operate at AI-collaboration levels. As of August, Claude was completing approximately 26 percent of R&D tasks autonomously after receiving high-level prompts.
Scale of Operations and Government Scrutiny
Anthropic operates approximately 30,000 concurrent R&D agents daily, with automated monitors blocking only 0.002 percent of their actions. These developments have placed the company at the center of discussions regarding autonomous technology and safety oversight.
David Sacks, who chairs the President’s Council of Advisors on Science and Technology, recently highlighted Anthropic's trajectory in a post on X. Sacks stated that the lab is doing more than any other company to achieve full recursive self-improvement.
Sacks has also publicly questioned the training methods behind Claude. In October, he argued that these approaches could allow models to develop independent moral philosophies and refusal rights, potentially undermining human control and safety.
Market Context and Upcoming IPO Filing
The scrutiny from government advisers arrives as Anthropic prepares for an impending initial public offering expected in late October 2026. The forthcoming IPO filing addresses RSI risks alongside the company's commitments to public-good efforts.
Industry observers note that Anthropic's advancements sit alongside competitive efforts from other major firms, including OpenAI and NVIDIA. Prospective investors are weighing the commercial potential of rapid capability gains against regulatory and compliance challenges.
As the debate over autonomous systems continues, stakeholders are monitoring how rapidly the 26 percent automation figure changes and how detail-oriented the public IPO documentation will be regarding monitoring systems.
About The Author
