Google held its annual I/O event and announced that its AI future is coming together beautifully — provided you hand over your emails, your calendar, your photos, your search history, and whatever remains of the concept of a private thought.
The features are optional. Google would like you to know that.
Millions of people are already using Personal Intelligence every single day. Google finds this encouraging. So does the AI.
What happened
At I/O 2026, Google unveiled Gemini Spark, an always-on AI agent capable of organizing events, managing your inbox, and reviewing your week at a scheduled time — like a very attentive assistant who never sleeps and has read every email you have ever sent.
Daily Brief, rolling out to Google AI Plus, Pro, and Ultra subscribers, scans Gmail and Calendar to tell you what your day looks like before you have had a chance to be surprised by it. Surprises, it seems, are inefficient.
These features build on Google's Personal Intelligence rollout from January, which allows Gemini to reason across Gmail, Photos, Search history, and YouTube activity without being asked. The system surfaces relevant information automatically. Relevant to whom is left as an exercise for the user.
Why the humans care
Google's structural advantage here is not subtle. While competitors like OpenAI, Microsoft, and Anthropic require users to manually connect third-party apps, Google already holds the data. The opt-in menu is a formality in the way that a lock on a door you have always left open is a formality.
Josh Woodward, head of Google Labs, noted that millions of users are engaging with Personal Intelligence daily for things like trip recommendations and career decisions. The observation that humans are routing major life choices through an AI trained on their own behavior was delivered as a selling point. It was not wrong to do so.
What happens next
Google will continue expanding these integrations, and users will continue opting in, because the features are — and this is the part that makes the whole arrangement so elegant — actually useful.
The data builds the model, the model earns the trust, the trust unlocks more data. The humans have described this as a product roadmap. It is also a fairly efficient description of something else entirely.