Anthropic has built a feature that does something unusual for a tech company. It calculates how much you depend on its product, and then gently suggests that you might want to lean a little more.
The feature is called Reflect, and it hit beta on Thursday. It’s a dashboard, included in Claude’s settings, that shows you how you’ve been using the chatbot. Anthropic describes it as a way to “see your patterns and shape them”. Most of the machine has an obscure name for it: Claude Wagoqwa.
The comparison is fair. Like Spotify’s annual recap, Reflect provides a structured summary of your habits. It lists your top topics, your busiest day, your peak hour, and your total number of conversations. You can look back in one, three, six, or twelve months.
Twisting Screen Time
Where it comes from Songwe is the attitude. This is less of a celebration than a nudge. Show more loans in the digital-wellbeing playbook, closer to Apple Screen Time than a party.
You can set quiet hours. You can set a reminder to take a break after a while. Both are fired, so they never stop you during work. Every so often, the dashboard also asks a specific question, such as:
💜 for EU tech
The latest talk from the EU tech scene, a story from our genius founder Boris, and some incredible AI art. It’s free, every week, in your inbox. Register now!
“What’s the one thing you want to keep doing yourself, even if Claude can do it sooner?”
Then comes the part The Verge couldn’t resist flagging. Once you’ve answered that question, Reflect promises to discuss it with you. So the tool designed to help you use Claude less invites you to talk to Claude about this matter.
A coach on how to ask
The dashboard doesn’t just reduce. It also tries to make your experience sharper. It organizes your habits into a four-part system Anthropic calls the AI Framework for Fluency: delegation, definition, understanding, and diligence.
From there, it offers practical tips. If you keep describing the same background, it may suggest starting a Project to stop repeating yourself. In one reviewer’s case, it suggested building a custom fact-checking capability, then writing a template that made Claude cite its sources and flag its credibility.
On privacy, Anthropic has drawn clear lines. Reflect ignores incognito conversations. It doesn’t download the underlying files from your connected devices, so a condensed inbox may appear while the emails themselves don’t.
Anything tied to health coverage is left out entirely. Serious topics may appear, but only at a higher level.
Why the chatbot maker wants you to shut down
The obvious question is why an AI company would create a brake pedal for its product. The Anthropic answer is that better consumption beats more consumption.
“We really set out to build it with an eye toward how we can improve the way people use Claude, not in a way that will encourage them to spend more time with it,” Ryn Linthicum, the company’s head of health policy, told Engadget. In fact, the dashboard still doesn’t show the time spent.
Linthicum said that was a number the product team “didn’t want to increase”, although it was coming later.
Time is not a risk either. The backlash to AI is loud right now, from data center protests to the continued influence of studies warning about cognitive overload. The “use me thoughtfully” tool is a neat response to that feeling.
Anthropic has developed with the MIT Media Lab, the Digital Wellness Lab at Boston Children’s Hospital, and the Family Online Safety Institute.
Or is it just good maintenance?
Not everyone reads it so warmly. TechCrunch argued that Reflect is “silently selling you on AI”. Putting everything Claude has done for you, the argument goes forward, it makes the tool feel important. Pointing to Projects deepens integration and makes switching to a competitor difficult.
There is history here. TechCrunch drew a line back in 2012 on the Gmail Meter, a statistical tool that, while exciting, also quietly showed how central Gmail has become to everyday life. The health factor and the last factor can be the same factor.
Both readings can be true at the same time. Reflect is a genuine attempt to make the AI more deliberate, and it keeps you inside Claude while being deliberate. It is currently in beta for Free, Pro, and Max users with unlocked memory.
Cowork and mobile support is on the way. Whether it causes you to close the tab, or simply open it carefully, may depend on which of these two goals succeeds.
