API
< Back to situation

[REVISION HISTORY]

ChatGPT functional capabilities and privacy features

1 recorded revision of this situation.

What changed

2026-11-12 23:32 UTC → 2026-11-14 03:00 UTC · added removed

Recent developments regarding ChatGPT highlight advancements in its functional capabilities and the availability of user privacy controls. The AI has demonstrated “Computer Use” abilities, allowing it to navigate macOS interfaces to perform complex tasks, such as creating custom shortcuts for file uploads. This functionality involves the AI testing its own work through system interactions and screenshots. Simultaneously, research into visual prompts has shown that the model can generate detailed visual representations of a user’s personal history, interests, and professional life based on accumulated conversation data. To manage such data, OpenAI provides several privacy features, including the ability to archive chats, use “Locked chats” for authentication, or utilize a “Temporary Chat” mode that prevents conversations from being saved to history. OpenAI has further clarified its data retention policies. Regular and archived chats remain in user accounts until manually deleted or removed by workspace policies. Once a chat is deleted, OpenAI schedules permanent deletion within 30 days, though exceptions may apply for legal, security, or de-identification purposes. While Temporary Chats are not saved to history, OpenAI may retain copies for up to 30 days for safety reasons. Additionally, disabling model improvement settings does not result in the deletion of conversations.

Versions

  1. 2026-11-14 03:00 UTC ChatGPT functional capabilities and privacy features
  2. 2026-11-12 23:32 UTC ChatGPT functional capabilities and privacy features

Select a version to see what changed compared to the one before it.