Artificial intelligence has become part of the daily lives of millions of users. Writing an email, preparing a presentation, summarizing a document, or coming up with an idea has become second nature for many. But at what point does a simple tool become a genuine habit? With Reflect, Anthropic is taking a novel approach. Rather than adding yet another feature to Claude, the company is launching a dashboard that analyzes how each person uses their assistant, to help users better understand their relationship with artificial intelligence.1
Going against the grain of most digital platforms, which seek to maximize the time spent on their services, Anthropic says it wants to encourage more mindful use of AI. The initiative opens up a new debate: after striving to make assistants ever more powerful, should companies now help their users better control their own usage?
Reflect turns your history into a personal dashboard
Reflect works like a retrospective of your activity with Claude. The tool analyzes your conversation history and builds a set of indicators that let you visualize how your usage has evolved over different periods, ranging from one month to a full year.1
Users can identify the periods when they rely on Claude most often, see the number of conversations they have started, and look at the types of tasks most frequently handed off to the assistant. Writing, brainstorming, programming, personal organization, or research: Reflect gradually highlights the areas where artificial intelligence plays the biggest role.
The goal is not to measure Claude’s performance, but rather to show how it fits into each user’s work and thinking habits.
A tool designed to prompt reflection rather than build loyalty
One of Reflect’s distinctive features is its approach. Whereas many digital services use dashboards to drive ever more engagement, Anthropic takes an almost opposite approach.
Beyond the charts and statistics, Reflect directly asks several questions designed to encourage personal reflection. In particular, users are invited to identify the tasks they would rather keep doing themselves, even if Claude could complete them faster. The aim is to encourage a more balanced use of artificial intelligence rather than systematic delegation.
Anthropic thus acknowledges that a powerful AI assistant does not necessarily need to be called upon for every daily activity. Some tasks retain a learning, creative, or reflective value that fully justifies doing them without assistance.
Reminders to restore a healthier balance
Reflect does more than observe usage habits. The tool also lets users schedule reminders that encourage them to take a break when they spend a lot of time with Claude.
It becomes possible to set time slots without using the assistant, or to receive notifications that simply invite you to step back for a moment. The feature is reminiscent of the screen-time tracking tools already available on some smartphones, but here it is applied directly to the use of artificial intelligence.
This direction is relatively unusual in the world of digital technology, where new services generally seek to increase the time users spend on them rather than limit it.
How to access Reflect
Reflect is currently available in beta to Claude Free, Pro and Max users who have enabled usage analytics.2
To access the dashboard, simply go to Claude’s settings, from the web app or the desktop version, and turn on the corresponding option. Once the feature is enabled, Reflect gradually generates a personalized report based on the available usage history.
If no report appears right away, Anthropic explains that this usually means not enough usage data has been collected yet, or that activity analysis has not yet been enabled on the account. The company also plans to extend Reflect to Cowork conversations soon, to provide a more complete view of collaborative uses of Claude.
The feature is gradually rolling out in countries where Claude is already available, including Europe, at no additional cost to users eligible for the beta.
A new approach to digital well-being
The launch of Reflect reflects an interesting shift in the artificial intelligence industry. Until now, competition between the major labs has focused mainly on model performance, context size, or reasoning capabilities.
Here, Anthropic is introducing another dimension: digital well-being. The company believes that the quality of an assistant depends not only on what it can produce, but also on how it supports its users over time.
This thinking echoes growing concerns in companies, educational institutions, and government agencies, where the rapid integration of AI assistants raises questions about preserving human skills, cognitive load, and gradual dependence on digital tools.
The ethical stakes: how much should we delegate to artificial intelligence?
Reflect also opens up a broader debate about the place AI assistants now hold in our daily activities. As these tools become more powerful, the temptation to delegate a growing number of tasks to them increases.
This shift brings obvious productivity benefits, but it also raises questions about users’ ability to retain certain essential skills, whether in writing, critical thinking, organization, or creativity.
By inviting everyone to think about the tasks they want to keep doing themselves, Anthropic is taking an approach that is relatively rare in the industry. The company implicitly acknowledges that the value of artificial intelligence lies not only in its ability to replace certain activities, but also in its capacity to support users without diminishing their autonomy.
Reflect marks a new stage in the evolution of AI assistants
With Reflect, Anthropic is not simply launching a new analytics tool. The company is offering a new way of interacting with artificial intelligence, in which transparency about usage becomes as important as the performance of the model itself.
As conversational assistants become embedded in professional and personal environments, this type of feature could gradually become standard. Understanding how we use artificial intelligence may soon be as important as the answers it gives us.
How does Reflect work?
Reflect is a new feature developed by Anthropic to help Claude users make more mindful use of artificial intelligence. Unlike conventional dashboards that merely display activity statistics, Reflect analyzes Claude usage habits to help each person understand the place the assistant holds in their daily life. The goal is not to evaluate the AI’s performance, but to provide an overall view of how it is used, in a spirit of transparency and digital well-being.
The system draws on the user’s conversation history to identify different usage trends. Reflect groups exchanges by period, highlights the times when Claude is used most, and classifies conversations by their main purpose: writing, programming, brainstorming, personal organization, learning, or information research. These analyses make it easier to see the habits that gradually take shape over weeks or months.
Beyond the statistics, Reflect introduces a reflective dimension. The feature regularly invites users to question how they use Claude, in particular by identifying the tasks they prefer to keep doing themselves and those they now delegate to artificial intelligence. This approach aims to encourage balanced use of AI rather than the systematic automation of intellectual work.
Reflect also lets users set up personalized reminders that prompt them to take a break or limit certain periods of use. The idea is inspired by the screen-time tracking tools found on smartphones, but this time applied to conversational assistants. Anthropic is thus seeking to promote responsible use of Claude rather than maximizing the time spent on its platform.
- Personalized dashboard: view your Claude activity over different periods (month, quarter, or year)
- Habit analysis: identification of the main types of tasks entrusted to artificial intelligence
- Usage statistics: tracking of the number of conversations, periods of activity, and how they evolve
- Guided reflection: questions designed to encourage more mindful use of AI
- Personalized reminders: option to schedule notifications prompting you to take a break or limit certain periods of use
- Native integration with Claude: works directly from the assistant’s settings, with no additional installation
- Gradual rollout: available in beta to Free, Pro and Max users who have enabled activity analysis
- Reflect is currently offered in beta, and some features are still being rolled out
- The analysis depends on the available history, so new users have little data at first
- Cowork conversations are not yet fully supported, but Anthropic has announced their integration
- The analyses are based solely on exchanges with Claude and do not reflect the user’s overall digital usage
- The statistics are a tool to support reflection and cannot be used to assess any potential psychological dependence on AI
- Users can still turn off the collection of the data Reflect needs to function
Learn more
Reflect is part of a broader effort by Anthropic to better understand how people actually use its models. On a related topic, read our article “Anthropic releases a major study: what 81,000 users really expect from AI”, which analyzes the expectations, uses and concerns expressed by Claude users.
References
1. Anthropic. (2026). Introducing Reflect for Claude.
https://www.anthropic.com/news/reflect-with-claude
2. Anthropic Support. (2026). Using Reflect in Claude.
https://support.anthropic.com