← Reddit

Please Claude, add a privacy mode

Reddit · lamedope · June 8, 2026
An individual experienced nervousness about screen sharing at work and with friends because sensitive browsing history was visible in the side panel. The person developed a concept design proposing a privacy solution to conceal certain history items during screen-sharing sessions.

Detailed Analysis

A Reddit user has surfaced a common usability concern with Claude's interface: the persistent visibility of conversation history in the side panel during screen sharing or in-person device viewing. The post describes discomfort experienced in both professional and social settings when others can see prior chat topics displayed in the sidebar, and proposes a conceptual UI solution — a "privacy mode" — to address this. While the article is brief and lacks technical depth, it represents direct user feedback about a genuine gap in Claude's current interface design.

The core issue is one of contextual privacy rather than data security. Claude's conversation history sidebar, which lists prior chat titles or topics, functions as an inadvertent disclosure mechanism when a screen is visible to others. Unlike a security vulnerability, this is a social and situational privacy concern: users may have queried Claude about sensitive personal matters, health questions, relationship issues, or work-related topics they do not wish colleagues or friends to observe. The discomfort described mirrors similar complaints historically raised about browser history and search autocomplete features, suggesting the problem is well-understood in UX design circles even if not yet resolved in Claude's interface.

The feature request connects to broader trends in AI assistant design, where the shift from ephemeral, session-based interactions to persistent, history-rich experiences creates new privacy surface areas. Competitors like ChatGPT and Google Gemini have similarly faced user requests for incognito or temporary modes, and ChatGPT has implemented a "temporary chat" feature that does not save history. The demand signals that as AI assistants become more deeply integrated into daily life, users expect the same contextual privacy controls they have grown accustomed to in browsers, messaging apps, and other personal tools.

Anthropic has not publicly announced a dedicated privacy or incognito mode for Claude as of mid-2026, making this an open product gap. The concept design shared in the post — though only viewable via a Reddit video link — suggests the user community is actively thinking through interface-level solutions rather than simply waiting for official features. Such grassroots design proposals often serve as valuable signal for product teams, indicating where user experience friction exists and what kinds of solutions users consider intuitive. The post's resonance within its community would likely reflect a latent demand that spans professional users, casual users, and anyone who shares devices or screens in the course of everyday life.

Read original article →