Claude Reflect is a beta feature inside Claude, the AI assistant built by Anthropic, that turns a person’s own chat history into a private summary of how they actually use AI: which topics come up most, what kinds of tasks eat the most time, and how that pattern shifts over a chosen stretch of time. It answers a question most AI usage dashboards have skipped — not “how many messages did I send,” but “what am I actually using this for, and is that serving me well?”
What it is
Reflect is not a productivity tracker in the workplace-monitoring sense, and it is not a billing dashboard. It is a self-review tool: a person opens it to see their own patterns, not for anyone else to see. Anthropic built it after user interviews turned up a recurring request — people wanted a clearer picture of how AI had actually worked its way into their day, beyond a vague sense that they “use it a lot.”
How it works
Reflect is available to Free, Pro, and Max plan users who have Claude’s memory feature turned on, since it draws on the same conversation history memory already uses. It lives in Settings on Claude’s web and desktop apps, and a person can review activity spanning the last 1, 3, 6, or 12 months. The summary covers recurring topics, usage patterns, and the types of tasks handled most often, and it periodically surfaces reflection prompts — questions like which task someone wants to keep doing themselves even if Claude could do it faster. It also supports quiet hours and break reminders after extended sessions.
Several categories are deliberately left out: incognito chats, the contents of files pulled in through connected tools, and health-related conversations are all excluded from the summary. The feature also draws on Anthropic’s own AI Fluency framework, a four-part model of delegation, description, discernment, and diligence developed with academic partners, and Reflect was built in consultation with digital wellbeing researchers from MIT Media Lab’s Advancing Humans with AI program.
Why it matters
As AI assistants move from occasional tool to daily habit, a gap has opened between how much people use these systems and how much they understand about their own use. Usage statistics that AI companies publish tend to describe the aggregate — how many people use a product, or how many tasks it handles — not what any one person is actually doing with it or whether that pattern is one they’d choose on purpose. Reflect is an early example of AI companies building that kind of self-awareness tool directly into the product, rather than leaving it to third-party screen-time apps that can’t see inside a chat history. Whether it changes behavior depends on whether people actually open it and act on what they see, but it does give a concrete, private answer to a question that’s otherwise easy to only guess at.
In the news
Anthropic introduced Reflect on July 9, 2026; see our report on the Claude Reflect launch for the announcement details.
FAQ
Does Claude Reflect cost extra? No. It’s included for Free, Pro, and Max users who have memory turned on — there’s no separate fee.
Can anyone else see my Reflect summary? No. It’s a personal view built from a person’s own conversation history, not a shared or admin-visible report.
Does it look at my private files or health conversations? No — Anthropic excludes incognito chats, the contents of files brought in via connected tools, and health-related conversations from the summary.
How do I turn it on? Turn on memory in Claude’s Settings, then open Reflect from the same Settings menu on the web or desktop app.
Sources: Anthropic — Reflect with Claude.