Google Now Lets Other AI Chatbots Control Your Smart Home

Google quietly opened up its smart home to rival AI on September 16, letting outside assistants — including Anthropic's Claude — look at and control your connected devices. Until now, Google Home answered only to Google's own assistant, Gemini. This new bridge means a different company's AI can check which lights are on, see whether your doors and windows are secure, review what happened while you were away, and switch things off. Think of it as giving a second house-sitter a key.

Why would you want that? Because most people already have a favorite AI they talk to all day. If that AI can see your home, you stop opening a separate app to ask 'did I leave the garage open?' You just ask the assistant you were already using. It's one less thing to juggle, and it's the direction everything is heading — AI that doesn't just answer questions but actually does things.

The catch is real, though. This isn't a checkbox in the Google Home app. You need a Premium Advanced subscription, a Google Cloud account, technical credentials, and a compatible AI app — a setup most people will find fiddly. Google also blocks the scariest action outright: AI agents cannot unlock your doors. And Google warns these assistants can behave unexpectedly, which is a polite way of saying a confused AI might do something you didn't intend. You can revoke access anytime.

One more thing worth knowing: sensitive data gets an extra lock. If you use Nest cameras with familiar-face recognition — the feature that tells you 'that's your neighbor, not a stranger' — a second person with access to your home setup has to approve sharing that separately. And AI can't yet create your automated routines, like 'lights off at 11pm.' That's coming later. Expect this to spread fast.