About AI Down Today

AI Down Today answers one question — is your AI down? — for people who just want to get their work done.

How the checking works

An automated monitor reads the official status feeds of OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini), and Microsoft (Copilot) every 75 seconds, around the clock. When a provider reports a user-facing incident, this site and the @aidowntoday Bluesky account say so in plain English: what's broken, since when, and what still works.

To keep the signal clean, an incident must persist across consecutive checks before we report it, and we only post state changes — new incident, escalation, all-clear. No engagement bait, no reposting rumors. Every page links to the provider's own status page so you can verify the source.

What this is not

We don't measure outages ourselves and don't crowdsource reports — we translate what the providers officially publish, faster and in plainer language than they do. If a provider hasn't acknowledged a problem yet, we won't show it yet either.

There is deliberately no AI in the reporting pipeline — a service that reports on AI outages shouldn't stop working during one.

Who runs it

Built and run by Joe Balewski — one human with one very reliable script. Joe works in media strategy and built this after watching one too many people assume a ChatGPT outage was their own fault. Not affiliated with OpenAI, Anthropic, Google, or Microsoft; nobody pays for placement here. Questions or corrections: reach out on Bluesky — corrections get fixed, not argued with.