Anthropic is adding a usage recap to Claude, giving users a dashboard that summarizes how they have been using the chatbot over the past month, three months, six months or year. The company announced the feature Thursday and calls it Reflect, because apparently even the analytics panel needs a wellness-adjacent name.
The feature turns Claude activity into a personal report: the subjects a user brings up, the kinds of work they hand off to the chatbot, and patterns such as the times of day they use it most. Anthropic said in a blog post that the dashboard is meant to help people see whether their Claude use matches their goals.
That makes Reflect part Spotify Wrapped, part screen-time monitor. The year-in-review format has already spread from music to products such as YouTube and Uber. Anthropic is applying the same basic mechanic to AI use, where the recap is less about flexing your most-played song and more about seeing how often you outsource thinking, writing, planning or whatever else you bring to Claude.
Anthropic said Reflect starts with a summary of a user’s main Claude topics, delegated task categories and usage habits. Users can also create “quiet hours” and set reminders to take breaks after defined amounts of time. The company said an analysis of total time spent using Claude is planned, which sounds like the missing number many people will either want immediately or regret seeing.
The dashboard is also designed to prod users with reflective prompts. Anthropic said Reflect may surface questions such as what a person wants to keep doing on their own, even when Claude could do it faster. After a user answers, Anthropic said the dashboard can offer a chance to discuss that answer with Claude.
That loop is the oddest part of the product: Claude will help you think about whether Claude is taking too much of your thinking. Anthropic has marketed Claude as an “AI collaborator” aimed at helping people think more carefully, and Reflect fits that pitch. It also makes plain that these tools are becoming measured environments, not just blank text boxes waiting for prompts.
Anthropic’s announcement, as described, does not add new model capabilities. It adds a mirror to the product: a way to inspect the metadata and themes around a person’s use. For users who rely on Claude heavily, that may be useful. For everyone else, it is another reminder that AI companies are not only building assistants. They are also building dashboards about the assistants.
This story draws on original reporting from The Verge.
