Anthropic to enable auto permission mode by default for Pro, Max and Team

On October 23, 2026, TestingCatalog reported on X that Anthropic will make auto mode the default for permission selection on its Pro, Max, and Team plans starting the following week. According to that account, Claude will no longer ask users to approve each step of an operation; instead it will rely on its internal safety evaluation. The report frames this as a shift in how permissions are handled on those subscription tiers rather than a change to the underlying models. The available information stops at this announcement: no specific rollout date beyond "next week" was given, and the material does not detail how broadly the mode applies, whether users can turn it off, or how Anthropic itself has described the change. No official statement from Anthropic is included in the source.