Anthropic enables Claude Code's automatic mode by default for all paid accounts

Anthropic is rolling out a significant shift in how Claude Code operates: the company will switch auto mode on by default for Pro, Max, and Team account holders beginning August 14, reducing the need for developers to approve each action the AI takes.

What auto mode does#

When enabled, auto mode allows Claude Code to execute programming tasks autonomously rather than pausing to request human sign-off at every step. The AI will proceed with actions freely unless Anthropic's systems identify them as potentially irreversible, destructive, or directed outside the user's environment—in which case it still asks for permission.

Safety claims outpace manual review#

Anthropic's own research suggests the automated approach is more reliable than relying on developers to review permission requests. In a trial involving over 1,000 paid users, auto mode flagged 89% of harmful actions, while human review caught only 13.6%. The company attributed this gap to decision fatigue: testers approved roughly 97% of all permission prompts, indicating that approval-based workflows can breed complacency.

Additional safeguards#

Beyond the mode switch, Anthropic has layered in new protective measures. The rollout includes prompt injection screening and custom hard deny rules—restrictions that teams can configure to block specific actions, such as unauthorized data transfers.

Adoption among internal teams#

Claude Code's head of product, Boris Cherny, noted that the internal team at Anthropic has been relying exclusively on auto mode for months, saying, "I couldn't imagine going back to permission prompts."