--- title: "Anthropic Launches Claude Reflect Beta to Track AI Habits" date: 2026-07-09 author: "Barry Elad" featured_image: "https://sqmagazine.co.uk/wp-content/uploads/2026/07/anthropic-launches-claude-reflect-beta-to-track-ai-habits.jpg" categories: - name: "Artificial Intelligence" url: "/artificial-intelligence.md" tags: - name: "News" url: "/tag/news.md" --- # Anthropic Launches Claude Reflect Beta to Track AI Habits Anthropic launched Reflect in beta on July 9, 2026, a Claude dashboard that shows users how they use the chatbot and lets them decide whether that time aligns with their goals. ## Quick Summary – TLDR: - Anthropic’s reflection dashboard can be found in Settings on Claude for web or the desktop app. - Users can choose one, three, six or **12**-month timeframes to review peak activity. - The tool scores habits against the 4D AI Fluency Framework rather than a simple time count. - Reflect excludes incognito chats and any conversation connected to a health integration tool from its insights. - The beta is available now for Free, Pro, and Max users who have Memory turned on, with Cowork reflections coming soon. ## What Happened? **Reflect**, per Anthropic, was built after interviews with users surfaced a common theme: a desire to better understand how, exactly, AI can be integrated into daily life. For readers already tracking their own screen time or wary of AI dependency, that framing is the point of difference: Reflect doesn’t block usage, it shows what a person actually did with Claude and leaves the judgment call to them. > Introducing a new way to reflect on how you use Claude. > > Your monthly recap shows when you use Claude most and what you spent that time working on, with options to set quiet hours and nudges to take breaks. Find your dashboard in Settings under Reflect: [pic.twitter.com/WzA3JfONlL](https://t.co/WzA3JfONlL) > > — Claude (@claudeai) [July 9, 2026](https://x.com/claudeai/status/2075225957711929667?ref_src=twsrc%5Etfw) ## Patterns, Quiet Hours, and a Fluency Score The dashboard opens with a summary of how a person has been using [Claude](https://sqmagazine.co.uk/claude-ai-statistics/), covering key topics, usage patterns, and the types of tasks they often work through. It also breaks down overall usage, peak activity, conversation line charts, and task breakdown charts for the selected window. Users can set quiet hours or schedule a nudge to take a break after a set amount of time using Claude. Both are reminders of the user’s own preferences, and either can be dismissed. Anthropic says it will soon add a specific view of total time spent using Claude, a metric the beta does not yet surface directly. ## What the 4D Fluency Score Actually Measures? The scoring system is not a simple usage counter; it grades behavior against four named skills tied to how a person works with Claude. Delegation covers setting goals and deciding whether and how to engage with AI. Description covers effectively describing goals to [prompt useful AI behaviors](https://sqmagazine.co.uk/prompt-engineering-statistics/) and outputs. Discernment covers accurately assessing the usefulness of AI outputs and behaviors, and Diligence covers taking responsibility for what a person does with AI and how they do it. A heavy user who delegates deliberately scores differently than one who copy-pastes without review, at identical hours logged. That distinction is the actual product: a self-assessment, not a bare screen-time tally. ## Privacy Boundaries: What Reflect Leaves Out? If a user asks Claude to summarize an inbox, the summary may appear in the reflection, but the source emails will not. Any conversation connected to a health integration tool is left out of the insights entirely. Anthropic worked with digital media and wellbeing experts from the **MIT Media Lab’s Advancing Humans** with AI program, the **Digital Wellness Lab at Boston Children’s Hospital**, and the Family Online Safety Institute to shape those boundaries. ## SQ Magazine’s Takeaway Screen-time counters are standard on phones and social apps; a comparable layer inside a chatbot’s own settings is new ground, and whether OpenAI or Google follow is an open question. The exclusion list Anthropic built in doubles as an acknowledgment that AI chat logs already function as informal health and productivity records, and folding those conversations into a habit dashboard risks creating exactly the sensitive profile users and regulators distrust. Routing the design past outside wellbeing researchers gives that boundary a documented reference point rather than a marketing claim. This launch reframes heavy Claude use as skilled use rather than overuse, provided the Discernment and Diligence scores hold up beyond launch. Cowork conversation reflections are coming soon, and the Memory requirement means casual users who keep Memory turned off won’t see a report until they opt in. Whether Reflect changes daily habits or becomes a screen-time widget people check once and ignore depends on whether Anthropic keeps updating it past this beta.