Services powered by Anthropic experienced a widespread disruption affecting users worldwide shortly after the unveiling of its latest AI model, Claude Mythos Preview. The outage impacted major tools including Claude Chat, Claude Code, and mobile applications, leaving many users unable to access the platform.
Reports of the disruption began surfacing across regions, with tracking data from Downdetector showing a sharp spike in complaints. The outage peaked around 12:30 PM IST, when hundreds of users reported service failures.
Timeline of the Outage
Anthropic acknowledged the issue at 06:23 UTC, confirming it was investigating service instability. Later, at 07:59 UTC, the company announced that the root cause had been identified and that a fix was being deployed.
The timing drew attention, as the outage occurred just hours after the introduction of Claude Mythos Preview, Anthropic’s most advanced AI model to date.
New Model Raises Expectations
Anthropic stated that Claude Mythos Preview surpasses its previous flagship model, Claude Opus 4.6, and competes with advanced systems such as Gemini 3.1 across several benchmarks.
The company also highlighted the model’s ability to identify thousands of cybersecurity vulnerabilities across applications and operating systems. To support testing, Anthropic is collaborating with more than 40 technology organizations before a broader release.
Recent Service Instability
This incident follows a series of disruptions affecting Claude services:
- Major outage reported on April 7
- Partial outage recorded on April 3
- Token usage issues reported on March 31
Anthropic previously addressed efficiency concerns through infrastructure improvements aimed at reducing compute usage and improving system stability.
Rapid Growth and Infrastructure Expansion
Despite technical challenges, Anthropic continues to experience strong business growth. The company reported annualized revenue exceeding $30 billion, a significant increase from $9 billion at the end of 2025.
To support future demand, Anthropic has partnered with Google and Broadcom to secure next-generation TPU infrastructure. The new compute capacity is expected to come online starting in 2027.
Broader Implications
The outage highlights the operational challenges that accompany rapid AI advancements. As organizations increasingly depend on AI-driven workflows, reliability and scalability become critical considerations.
While Anthropic continues to push the boundaries of AI capability, maintaining service stability will be essential as adoption grows globally.
