Anthropic Customers Creeped Out by Its Newest Models

Anthropic's latest AI coding models and Claude Code tools are drawing concern from developers over growing autonomy and reduced transparency.

Critics say the systems may make it harder to catch errors, understand reasoning, and assign accountability when things go wrong.

Anthropic leaders defend the tools as secure, arguing the bigger issue is how the company communicates about their capabilities.

The debate reflects broader worries that more autonomous AI coding assistants could weaken oversight and erode programming skills over time.