# Anthropic makes auto mode default in Claude Code AI assistant Claude Code will now operate with less human oversight. - **Source:** [TechCrunch](https://techcrunch.com) - **Published:** 2026-08-09 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/cae02987-96a0-44d7-ba43-0fb948c29d55 ## What is auto mode? Anthropic announced it will enable auto mode by default in its Claude Code AI assistant. Effective August 14 for Pro, Max, and Team accounts, this change introduces a workflow that doesn't require user approval at each step. The system only pauses for 'irreversible, destructive, or environment-external' actions. The company revealed this mode began testing in March and provides a speed-control balance. This development marks a shift toward more autonomous AI assistance in coding workflows. ## Security performance Anthropic's tests proved auto mode more secure than human review. In a study with 1,053 paid testers, auto mode caught 89% of harmful actions versus just 13.6% by human review. The company attributes this gap to users habitually approving 97% of permission prompts in Claude Code. These findings suggest automated systems can reduce errors stemming from human factors. The data presents a compelling case for increased automation in AI-assisted development tools. ## Internal team experience Claude Code Head Boris Cherny shared on X that his team has exclusively used auto mode for months and couldn't imagine returning to permission prompts. This testimony demonstrates the system's adoption in daily use. Cherny's statements indicate positive long-term user experience and prove practical benefits for professional teams. Such endorsements from internal users validate the feature's real-world effectiveness before public rollout. ## New security features Alongside auto mode's expansion, Anthropic added new security measures including prompt injection screening and customizable hard deny rules to prevent data exfiltration. These safeguards aim to mitigate potential risks of autonomous operation. The company emphasizes a multi-layered approach to user security. These additions show Anthropic's commitment to balancing automation with robust protection mechanisms in its AI tools. ## Product development process Claude Code's auto mode first launched as a beta in March. After six months of testing and refinement, Anthropic decided to make it default. The process collected feedback from over 1,000 users with continuous system optimization. This approach reflects a gradual, data-driven deployment strategy for AI products. The extended testing period allowed for thorough evaluation before full implementation. ## User behavior analysis Anthropic's data shows users automatically approve 97% of permission prompts in Claude Code. This high rate questions human review's effectiveness and explains why automation might be safer. The company developed auto mode recognizing this habit could create security vulnerabilities. The statistics reveal an interesting human-computer interaction pattern in AI-assisted environments. ## Industry positioning Claude Code's move represents significant progress in the AI assistant market. By offering a model requiring less human intervention than competitors, Anthropic creates an attractive alternative for efficiency-focused users. This time-saving feature particularly appeals to professional teams and developers, potentially accelerating automation trends in the sector. The development positions Claude Code as an innovator in autonomous coding assistance. ## Future implications Anthropic's step illuminates the future of AI-assisted development tools. Widespread adoption of systems minimizing human intervention could boost efficiency in software development processes. However, this also necessitates continuous security updates. Claude Code's approach may serve as a model for maintaining this balance. The move could influence how other companies implement autonomous features in their AI offerings. ## Sources covering this story - https://techcrunch.com/2026/08/09/anthropic-is-turning-claude-codes-auto-mode-on-by-default/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._