Anthropic AI Updates: August 11, 2026

1. Claude Advances a Bound Related to the Riemann Hypothesis

Anthropic. Anthropic reported that an unreleased research version of Claude improved a longstanding bound tied to the Riemann hypothesis, raising the proven proportion of Riemann zeta function zeros satisfying the hypothesis from 41.6% to 67.2%. Working across two sessions in Claude Code, the model generated 650 initial ideas, coordinated roughly 60 subagents, ran about 2,400 shell commands and thousands of numerical checks, and produced a formally verifiable proof in Lean that human mathematicians at Anthropic verified. Anthropic cautioned that it does not expect these techniques to lead to a full proof of the Riemann hypothesis, but the result shows how agentic tool use and subagent coordination can extend a model into original mathematical research. Source

2. Anthropic Makes Claude Code’s Auto Mode the Default

Anthropic. Anthropic is switching Claude Code’s auto mode on by default, letting the agent plan and execute multi-step coding work with less step-by-step human approval. The change pushes more developers toward hands-off agentic workflows and follows a broader industry move to reduce the friction of confirming each tool call. For teams, it raises the stakes on sandboxing and permission configuration, since a more autonomous default means the agent acts before a human reviews each step. Source