Anthropic's Reflection: AI gets its screen-time moment
Add Axios as your preferred source to
see more of our stories on Google.

A screenshot of Claude's new Reflection feature. Image: Anthropic
Anthropic is testing a new Reflection tool that helps Claude users see when they use the chatbot, what they use it for and whether that lines up with their goals.
Why it matters: AI chatbots are becoming work tool, search engine, confidant and coach all at once — raising the same kinds of questions about healthy use that social media did, but with even more intimate data.
Driving the news: Announced Thursday, Reflection lets users look back over one, three, six or 12 months of Claude activity, including key topics, usage patterns and the kinds of tasks they use Claude for.
- It also offers prompts about AI's role in their lives, plus optional quiet hours and break reminders.
- The tool is being released as a beta and is available to free and paid users who have Claude's memory feature turned on.
What they're saying: "In our interviews with users, a common theme that's emerged is a desire to better understand how AI could be integrated into daily life," Anthropic said in a blog post.
- "How often should someone use AI? How can it be used most effectively? When is AI suited to a task, and when is it better left to a human?" it said. "We built this feature to help answer these types of questions."
The intrigue: The feature taps Claude to help users decide when not to use Claude.
Between the lines: Anthropic said only certain types of data and usage are captured in Reflections.
- It doesn't draw from incognito chats, nor from any data generated by connected health tools.
- With other connected tools, it is more complicated. Anthropic says that, for example, if you ask Claude to summarize your inbox, the summary might show up in a reflection, but not the underlying emails.
Yes, but: Even high-level summaries can reveal sensitive personal patterns, the company acknowledges.
- "We built this tool to be a reflection of how you use Claude," Anthropic said in a blog post. "For some users, this may include sensitive or more personal conversations.
