Skip to content
AI ConnectPowered by VELENTIS
AI-generated1 min

OpenAI Launches Computer History Feature for macOS Desktop App

OpenAI rolls out Computer History on macOS to give ChatGPT contextual workflow memory, shifting away from screenshots while withholding the feature in the EEA.

This article was AI-generated and published automatically. Context, labelling and all sources at the end of the article.

(KI-generiertes Symbolbild: Gemini / AI Connect)

OpenAI introduced a new capability called Computer History for its macOS desktop application on August 13, 2026. The functionality is available on an opt-in basis exclusively to subscribers on ChatGPT Pro, Business, and Enterprise plans. The feature is designed to provide the AI assistant with persistent background context regarding user workflows across different desktop tasks.

The system formally replaces an earlier experimental prototype known as Chronicle, which relied on continuous screen recording and periodic screenshots. Computer History abandons visual screen capture altogether, instead interfacing directly with the native macOS accessibility framework. This allows the application to capture discrete operating system events, such as mouse clicks, keyboard input, and application transitions.

Rather than streaming raw telemetry to remote infrastructure, the captured interactions are processed directly on the user device. The software condenses individual user actions into structured, local Markdown summaries that ChatGPT can reference during subsequent conversations. This architecture reduces computational overhead while giving users direct visibility over the generated contextual records.

Despite the shift toward structured text aggregation, OpenAI has implemented strict geographic restrictions. The Computer History feature is currently unavailable across the European Economic Area, Switzerland, and the United Kingdom. The restriction stems from compliance requirements under strict European data protection standards, with no confirmed release date for those markets.

Alongside regulatory barriers, OpenAI explicitly highlighted potential security vulnerabilities inherent in system-wide activity logging. The primary concern centers on indirect prompt injection vectors, wherein malicious instructions embedded within displayed documents or websites could enter the aggregated history and manipulate model outputs during future interactions.

What this means for you

For desktop power users outside Europe, this architecture provides a pragmatic alternative to screen recording for workflow memory. However, the exclusion of European markets and lingering prompt-injection vulnerabilities mean enterprises must carefully weigh the security tradeoffs before enabling OS-level context logging.

Evidence

Solidly sourced
46/100
  • The tool replaces the screenshot-based Chronicle experiment by tracking discrete events via the macOS accessibility API.

    single source
  • Interaction events including clicks, keystrokes, and application switches are aggregated locally into Markdown summaries.

    single source
  • The rollout excludes the EEA, Switzerland, and the UK due to privacy rules, and OpenAI warns of indirect prompt injection risks.

    single source

The evidence score is computed, not hand-set: from confidence, the number of sources and the share of verified statements.

Source & transparency

As of: August 16, 2026

AI-generatedAI-generated: produced automatically from vetted sources with technical quality checks (source, quote and figure verification); no human sign-off of each item before publication

Sources
1
Verified statements
0 / 3
Evidence score
46Solidly sourced

Want to put this into practice?

We connect you with suitable AI providers from the DACH region, free of charge and without obligation.

What's next?