OpenAI's Computer History turns clicks into searchable ChatGPT memories

OpenAI has introduced Computer History, a macOS feature that transforms how users interact with ChatGPT by creating a searchable timeline of their digital activities. Unlike the earlier "Chronicle" research preview, which captured screenshots, Computer History records interaction events: clicks, keystrokes, keyboard shortcuts, and app switches, all read through macOS's accessibility system. Screenshots, screen recordings, microphone input, and system audio are explicitly not captured, and activity in private browsing mode is never included, according to OpenAI.

The system converts these captured events into text summaries at regular intervals, storing them locally as Markdown memory files. The timeline view groups these summaries by day and time and identifies which applications were involved. This design enables Computer History to detect recurring workflows—when ChatGPT identifies repeatable patterns in a user's activity, the timeline suggests a "Skill" or automation template that users can ask Codex to build. Typical use cases cited by OpenAI include finding recently edited documents or generating standup summaries from recorded activity.

Access to Computer History is gated by a multi-layered permission system. In Business and Enterprise workspaces, an admin must explicitly enable the feature first; individual users must then consent separately to have it activated. Users retain granular control over which apps and websites contribute data through include and exclude lists, and they can pause or resume recording at any time from the macOS menu bar. However, Computer History is not yet available in the European Economic Area, Switzerland, or the United Kingdom.

Data storage presents a notable security consideration. Temporary event files sit on the Mac within the ChatGPT app group and are deleted after 48 hours. The memory files generated from them, however, remain as plaintext Markdown on the local file system until the user manually deletes them—and critically, these files are not encrypted. This means any program running under the same macOS user account could potentially read them. OpenAI processes temporary event files on its servers during memory generation but states it does not retain them after processing unless legally required. The files are not used for AI training, OpenAI says, though chat contents that include memories as context may become training data depending on the user's data control settings. OpenAI's own documentation flags a heightened prompt injection risk: because Computer History captures content from apps and websites, malicious instructions embedded on a webpage could trick ChatGPT or Codex into following them. The company recommends pausing recording or excluding apps entirely when handling sensitive health, financial, or personal data, and advises against using the feature with communication apps unless all participants have explicitly consented.