Published on July 7, 2026 at 09:33 Comment 4 𝕏 Print © Anthropic Anthropic says it has discovered an unknown structure inside Claude. This architecture, which is very similar to that of the human mind, enables the AI model to reason. It has never been implemented by Anthropic researchers. The structure emerged on its own in Claude’s brain. For the startup, it is a form of “conscious access.” Claude, Anthropic’s generative AI, has just reached a milestone.