OpenAI, Anthropic and Google all reported service disruptions Thursday, hitting three of the world's most-used AI platforms within the same window.
Driving the news: OpenAI said it was investigating 'elevated errors across ChatGPT and Codex,' with the issue ongoing for at least 20 minutes at the time a status-page screenshot was reviewed.
Anthropic reported elevated errors across multiple Claude models starting just before 9:30 a.m. ET. According to Anthropic's status page, the affected models included Mythos/Fable 5.1, Mythos/Fable 5, Opus 5, Opus 4.8 and Opus 4.6. By 10:49 a.m., the company said it had identified the cause and was 'continuing to work on a fix.'
Google's Gemini was also hit. According to Google AI Studio's status page, the Gemini API was having trouble serving recently created API keys, including keys used through OpenAI-compatible libraries. Google said it was investigating.
Why it matters: Millions of consumers and businesses — from coders to customer-service bots — now run on infrastructure controlled by three private companies. When they stumble, so does everyone who built on top of them.
The bottom line: This is what concentrated market power looks like in practice: no elections, no regulators, just three dashboards deciding whether the digital economy functions on a given morning. Competition, not government mandates, is what will eventually force these platforms to build resilience — customers vote with their wallets when uptime fails.