On June 1, 2026, Anthropic's Claude development team announced that they had reset the 5-hour and weekly usage limits for all Pro and Max subscribers. This move was aimed at resolving a critical issue involving the Claude Code programming tool that had occurred shortly before.
What Happened
According to information from the official X account of Claude developers, a technical bug in Claude Code sessions caused the tool to repeatedly trigger too many unintended parallel subagents. As a result, system resources were consumed much faster than expected, causing many developers to unknowingly exhaust their rate limits in a short period of time.
Background
Claude Code is one of the latest command-line programming assistant tools favored by developers for its ability to automate code writing and editing through agent mechanisms. Running parallel subagents is a core feature that helps break down and solve complex programming tasks. However, the lack of control over the number of these subprocesses inadvertently turned a useful feature into a resource-hogging tool, directly affecting the experience of the platform's highest-paying customers.
Why It Matters
This incident once again highlights the risks of relying on AI-agent-based development tools that autonomously make decisions and execute background tasks. For the developer community in Vietnam, especially those integrating Claude Code into actual production workflows, Anthropic's prompt acknowledgment of the bug and refund of usage limits is a plus for customer service. Nonetheless, it serves as a reminder for the tech community to closely monitor the behavior of autonomous agents to avoid unforeseen and out-of-control resource costs.