Anthropic has updated its Claude Code product by setting Auto Mode as the default for Pro, Max, and Team plans starting August 14, aiming to enhance user safety. This change reflects the company's effort to better manage potentially harmful commands within its AI system.
According to The Decoder, the classifier integrated in Auto Mode detected 89 percent of dangerous commands, significantly outperforming human reviewers who identified only 13.6 percent. This improvement highlights the effectiveness of automated safety features in mitigating risks associated with AI-driven tools.
For Japanese investors and tech users, Anthropic's move underscores the growing importance of AI safety protocols, particularly as automation and AI integration continue expanding in financial and technological sectors across Asia.