ChatGPT macOS desktop app adds Computer History that logs clicks and keystrokes
OpenAI has rolled out Computer History in the ChatGPT desktop app for macOS, an opt-in capability that records clicks, keystrokes, and app switches to build a searchable timeline ChatGPT and Codex can reference in future conversations. Reporting frames the feature as a shift from stateless chat toward persistent, workflow-aware assistance that can suggest automations and resume incomplete tasks. OpenAI says users enable it under Settings → Integrations and that rollout is global for Pro and Business tiers, with ZDNet noting Enterprise eligibility; the company positions it as an evolution of its Chronicle research preview with added privacy controls. Third-party coverage highlights a material caveat: The Decoder reports local storage as unencrypted Markdown files, leaving open questions about training use and enterprise security.
ChatGPT macOS desktop app adds Computer History that logs clicks and keystrokes
ChatGPT's desktop app on macOS has a new feature called Computer History that turns your actions into training data. The feature learns how you work, suggests automations, and can pick up tasks you left half done.
Key takeaway
The boundary between conversational AI and an OS-level behavioral agent is narrowing as ChatGPT ingests click and keystroke timelines from the Mac desktop.
What happened
According to The Verge and OpenAI, ChatGPT's macOS desktop app now offers Computer History, a feature that logs user clicks and keystrokes to build a personal activity timeline used for personalized assistance and workflow automation suggestions.
The Decoder reports the timeline is stored locally as unencrypted Markdown files, while OpenAI's X posts state rollout is global for Pro and Business users who opt in via Settings → Integrations, building on the earlier Chronicle research preview.
Evidence
Computer History logs clicks and keystrokes on macOS to build a personal timeline for ChatGPT.
The Verge AI · attributed
ChatGPT's macOS desktop app introduces 'Computer History,' a feature that logs user clicks and keystrokes to build a personal timeline.
Computer History records clicks, keystrokes, and app switches and stores data locally as unencrypted Markdown files.
The Decoder · attributed
OpenAI's Computer History records clicks, keystrokes, and app switches on Mac and turns them into a searchable timeline for ChatGPT and Codex. The data is stored locally as unencrypted Markdown files.
Computer History is opt-in and rolling out globally to Pro and Business users on Mac.
OpenAI · attributed
To use Computer History, opt in under Settings → Integrations in the ChatGPT desktop app on Mac. Rolling out globally now to Pro, Business,
Pro, Business, and Enterprise users can opt in to app and website usage tracking integrated into conversations.
ZDNet AI · attributed
ChatGPT has introduced 'Computer History' for Mac, a feature that allows Pro, Business, and Enterprise users to opt in to tracking app and website usage to create a timeline.
Computer History builds on the Chronicle research preview with reduced token usage and more privacy controls.
OpenAI · attributed
Computer History builds on the Chronicle research preview with reduced token usage and more privacy controls. A new timeline view gives you
OpenAI launched Computer History as an opt-in macOS feature turning recent activity into memories for ChatGPT and Codex.
Techmeme · attributed
OpenAI launches Computer History, an opt-in feature that turns recent computer activity on macOS into memories and a timeline that ChatGPT and Codex can use
Why it matters
Builders designing assistant products must now plan for longitudinal behavior capture, explicit opt-in consent flows, and local-data security models rather than stateless chat context alone.
Limits and uncertainties
Available excerpts truncate OpenAI's statements on whether Computer History data is used for model training.
Reporting describes local timeline storage as unencrypted Markdown files but does not fully detail encryption, retention, or enterprise governance controls.
Practical implications
Teams shipping desktop agents should treat OS-level activity logging as a high-sensitivity consent and security surface, not a default-on convenience feature.
Product designers can reduce prompt friction by anchoring assistant memory in workflow timelines, but must audit local file storage and access boundaries.
What to watch
Whether OpenAI publishes clearer policy on training use, retention, and encryption for Computer History timelines.
How Pro, Business, and Enterprise customers respond to unencrypted local Markdown storage in security reviews.
