Anthropic introduces Reflect with Claude beta

Anthropic has introduced a beta feature called Reflect with Claude, designed to help users understand and optimize how they integrate AI into their daily routines. The tool provides a reflection dashboard that enables users to visualize usage patterns and align their AI interaction with their personal and professional goals.

Usage Tracking and Visualization

Reflect with Claude allows users to analyze their chat activity over selectable timeframes of 1, 3, 6, or 12 months. The dashboard provides a breakdown of when Claude is used most frequently and the specific topics and tasks the user has worked through.

Key capabilities of the reflection dashboard include:

  • Pattern Analysis: A summary of key topics and usage patterns.
  • Time Tracking: While currently focused on activity and topics, Anthropic plans to add a view showing the amount of time spent using Claude in the future.
  • Wellness Tools: Users can set "quiet hours" or schedule nudges to take breaks after a specific duration of use.
  • Interactive Reflection: The system periodically presents questions—such as asking what tasks a user prefers to keep doing themselves despite AI efficiency—to encourage users to discuss their AI role with Claude.

AI Fluency and Skill Development

Reflect with Claude integrates the 4D AI Fluency Framework to help users sharpen their collaboration with the AI. The tool analyzes user activity across four dimensions:

  1. Delegation: Deciding whether and how to engage with AI and setting goals.
  2. Description: Prompting useful AI behaviors and outputs through effective goal description.
  3. Discernment: Assessing the accuracy and usefulness of of AI outputs.
  4. Diligence: Taking responsibility for the actions taken with AI.

The reflection report provides summaries of these dimensions with specific examples of user behavior, such as noting if a user typically settles strategy before delegating tasks. It also offers practical suggestions, such as recommending the use of Projects to avoid repeating context for ongoing work.

Privacy and Safety Guardrails

To protect user privacy, Anthropic implemented specific exclusions for the reflection tool. Insights are not drawn from incognito chats or underlying files from connected tools; for example, if Claude summarizes an inbox, the summary is reflected, but the source emails are not. Additionally, all conversations connected to health integration tools are entirely excluded from the reflection insights.

Anthropic collaborated with digital media and wellbeing experts from the MIT Media Lab's Advancing Humans with AI (AHA) program, the Digital Wellness Lab at Boston Children’s Hospital, and the Family Online Safety Institute to handle sensitive or personal conversations. While sensitive topics may appear in the reflection, they are presented only at a high level.

Availability and Requirements

Reflect with Claude is currently in beta for Free, Pro, and Max users. To access the feature, users must have the Memory feature enabled. The tool is available via the Settings menu on the web and desktop applications. Support for reflecting on Cowork conversations is listed as coming soon.

Sources

Related

  • Dispatch
    Anthropic Independent Research Pilot on Claude Usage Data

    Anthropic has piloted a program allowing external researchers to analyze real-world Claude usage data via a privacy-preserving tool called Anthropic Insights, releasing aggregate findings on human-AI collaboration and productivity.

  • Dispatch
    Anthropic Clio privacy-preserving analysis tool

    Anthropic announced Clio, an automated, privacy‑preserving system that aggregates real‑world Claude usage into abstract topic clusters to improve safety monitoring without exposing user data.

  • Dispatch
    Anthropic Claude Tag launch

    Anthropic announced Claude Tag, a Slack‑integrated, team‑focused AI assistant that can be @‑mentioned to delegate tasks, learn context, and act proactively, now available in beta for Claude Enterprise and Team customers.

  • Dispatch
    Anthropic Introduces Claude Projects for Collaborative AI Workflows

    Anthropic has launched Projects for Claude Pro and Team users, allowing them to organize chats with curated knowledge sets and custom instructions within a 200K context window.

  • Dispatch
    Introducing Claude

    Anthropic has released Claude, a next-generation AI assistant designed to be helpful, honest, and harmless, available in high-performance and fast, lightweight versions.