Claude has suffered several outages in the past week
Getty Images
Anthropic has admitted that a surge in demand for its Claude chatbot has led to several service disruptions over the past week.
The AI company has been the beneficiary of a wave of people leaving ChatGPT, after OpenAI signed a controversial deal with the U.S. Department of Defense. A day after Anthropic refused to allow its AI to be used for mass surveillance of citizens and fully automated weapons systems, OpenAI agreed a deal to supply its AI models for “any legal purpose,” although OpenAI subsequently insisted it wouldn’t be used for surveillance of U.S. citizens.
Earlier this week, a boycott group estimated that around 1.5 million customers had pledged to leave ChatGPT, but that figure has subsequently risen to more than 2.5 million.
Claude Outages
Claude's service status page highlights several recent disruptions
Barry Collins
Claude has suffered a series of service disruptions over the past week, as it struggles to cope with the demand of an influx of new users.
The Claude.ai service suffered a partial outage lasting more than four hours on Feb. 27, another lasting almost three hours a day later and a further outage lasting for almost three hours on March 2. A much shorter 48-minute outage was recorded on March 3.
The service status page shows similar disruption to the Claude platform (formerly called console), the API used by developers and Claude Code.
Claude hit the top of the U.S. App Store charts on Feb. 28, overtaking both ChatGPT and Google Gemini as the most downloaded free app. This swift uptick in demand is responsible for the service wobbles, Anthropic claims.
“We’ve seen incredible demand for Claude over the last week and our team is doing everything we can to scale our infrastructure as quickly as possible to keep pace with the recent surge,” the company said in a statement. “We’re hugely appreciative of everyone’s understanding while we work through this.”
Anthropic Fights On
The Pentagon officially confirmed on Thursday that Anthropic would be deemed a supply-chain risk, threatening its ability to do business with military contractors or suppliers. Anthropic has pledged to challenge the decision in court, although the designation might not do as much damage to the company’s business as first expected.
For example, a Microsoft spokesperson told CNN that its lawyers had concluded that it could still offer Anthropic products, including Claude, to its customers “other than the Department of War,” the Trump administration’s new designation for the Department of Defense. However, another defense contractor, Lockheed Martin, said it was cutting ties with Anthropic.
OpenAI, meanwhile, has been attempting to quell the backlash against its deal with the U.S. government, putting out a blog post claiming that “our tools will not be used to conduct domestic surveillance of U.S. persons,” and that “our services will not be used by Department of War intelligence agencies like the NSA.”
The blog post also takes a dig at Anthropic, claiming the OpenAI contract has more guardrails than Anthropic’s original deal. “We don’t know why Anthropic could not reach this deal, and we hope that they and more labs will consider it,” the OpenAI statement reads.
OpenAI also states that it doesn’t believe Anthropic should be designated as a supply chain risk “and we have made our position on this clear to the government.”