Claude AI suffers widespread outage as bug drains user quotas within minutes

Claude AI Experiences Major Service Disruption

Claude AI went offline for thousands of users on June 2, 2026, marking a significant AI outage that exposed vulnerabilities in Anthropic's scaling infrastructure1

1

. The disruption began around 2:10 AM ET when reports started flooding Downdetector, eventually reaching 216 reports in the UK and 139 in the US1

1

. Anthropic confirmed the incident through its status page, acknowledging elevated error rates affecting multiple services and stating that engineers were implementing a fix3

3

.

The Claude service disruption extended across the entire ecosystem, impacting Claude AI, Claude API, Claude Console, and Claude Code3

3

. Users reported failed prompts, interrupted sessions, loading problems, and difficulties generating responses. The widespread access issues affected both casual users and developers who rely on the Anthropic AI chatbot for critical workflows.

Critical Bug in Claude Code Drains User Quotas

The root cause appears linked to a malfunction in Claude Code's sub-agent system, which turned a productivity feature into a resource-draining bug2

2

. Claude Code uses sub-agents to divide complicated tasks into smaller parallel processes, but the bug caused these sub-agents to multiply excessively, resulting in unusually high token consumption2

2

.

Paid Claude Pro and Claude Max subscribers bore the brunt of this technical failure. Some Max plan users reported that usage allowances designed to last several hours were depleted within minutes2

2

. Screenshots shared online showed account allowances dropping dramatically over short periods, creating what many described as a "silent drain" rather than a complete outage2

2

. Anthropic responded by rolling out an emergency quota reset for impacted accounts2

2

.

Broad Impact Across User Communities

The AI assistant partial outage affected diverse user groups who have integrated Claude into their daily operations. Software developers relying on Claude Code for development, debugging, and automation found their productivity feature had become a major source of unexpected quota consumption2

2

. Students using the AI chatbot for research, exam preparation, and learning assistance encountered slower responses or temporary service instability2

2

. Business teams operating on paid plans faced concerns over unexpectedly depleted usage allowances, potentially affecting ongoing projects and workflows.

As complaints mounted, users flocked to social media to share updates, screenshots, and error messages. The outage sparked discussions about how heavily people now rely on AI tools for work, coding, research, and content creation, with some expressing frustration over unexpected downtime during active projects3

3

.

Pattern of Performance Issues Raises Questions

This incident follows previous complaints from developers regarding rapid quota depletion and performance inconsistencies in Claude Code. Over recent months, some users have pointed to issues involving context management, agent behavior, and heavy system loads during peak demand periods. Anthropic has previously acknowledged challenges associated with scaling advanced AI workflows.

Industry observers note that advanced AI systems are increasingly incorporating features such as parallel agents and autonomous workflows. While these capabilities improve performance, they can also introduce new technical challenges when deployed at scale. The repeated issues suggest Anthropic may need to strengthen its testing protocols and monitoring systems to detect similar problems before they impact paying customers.

Until the platform returns to full stability, users are advised to monitor usage levels through the Claude dashboard, limit extensive parallel tasks and long coding sessions, use standard chat mode if Claude Code appears unstable, and keep alternative AI tools available for urgent work.