Subvolts take Manage AI agent permissions and data sharing across OpenAI tools for better control. Oct 19, 2026 This FAQ provides details on OpenAI's "Dots" feature, explaining its privacy, security, and safety aspects.
- Developers
- Understand how your plugins integrate with OpenAI's agent-like features and share permissions across their ecosystem.
- Business
- Learn how data usage for model training can be controlled, especially for enterprise workspaces where data is not used by default.
- Everyday users
- Grasp how your AI assistant uses tools, manages your data, and what controls you have over its access and learning.
This clarifies how OpenAI's AI agents, called Dots, integrate with existing tools and manage user data, impacting privacy and control.
Dots privacy, security, and safety FAQs
Learn what your dot can access, how it uses and protects your information, and when it needs your approval.
Yes. Plugin permissions are shared across dots, ChatGPT, ChatGPT Work, and Codex. Your dot can use your existing connections within the permissions you’ve granted.Your dot comes with strong defaults for which actions need your approval each time, which you can approve in advance, and which it can take without asking. See the Actions, approvals, and control section for more details.
You can connect new plugins and manage existing ones in ChatGPT’s Plugins tab. These connections are shared across dots, ChatGPT, ChatGPT Work, and Codex.
You can also set up a separate Slack account for your dot, giving it its own identity and access in those services. Your dot can still use the plugins you’ve already connected to ChatGPT, even if it has its own email or Slack account.
You can manage access to different plugins in ChatGPT’s Plugins tab. Your dot follows strong defaults about when it needs your approval before taking actions through plugins.Custom Rules let you set additional boundaries for dots, such as telling your dot never to send emails. These rules cannot override certain built-in guardrails and safety…
Opening of the original on OpenAI Help Center From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
OpenAI's "Dots" feature shares plugin permissions across ChatGPT, ChatGPT Work, and Codex. Users manage these connections and set custom rules for their dot's actions. Dots can access conversation context and memories from ChatGPT, with options to control data usage for model improvement. OpenAI encrypts data in transit and at rest.
Your AI assistant, called a "dot," can use tools you've already given permission to ChatGPT. You decide what it can do automatically and what it needs to ask you first. It can remember past chats, but you control if that information helps train the AI.
This update clarifies how "Dots," OpenAI's agent-like feature, integrates with existing ChatGPT functionalities and plugins. It emphasizes user control over permissions and actions, a key differentiator in the evolving AI agent landscape. By sharing plugin access across its ecosystem, OpenAI streamlines the user experience. The focus on granular control and data privacy settings addresses growing consumer concerns. Future developments will likely center on expanding proactive research capabilities and refining memory management.
- Plugin Permissions Shared Across: Dots, ChatGPT, ChatGPT Work, Codex
- Data Encryption: In transit and at rest
- Model Training Control: Via "Improve the model for everyone" setting
- Data Deletion: Delete dot to delete its context
- plugins
- Software components that add specific features to a larger application.
- Add-ons that give AI new abilities, like browsing the web or writing code.
- context
- Information provided to an AI model that helps it understand and respond appropriately to a query.
- The background information and previous conversation that helps the AI understand what you're talking about.
Go to ChatGPT's Plugins tab. Review your connected plugins and consider which ones you want your dot to access. You can also explore setting up a separate Slack account for your dot to manage its identity and access.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news. Pulled from this item and others like it.
- RuleSet custom rules for dot actions and permissionsDefine boundaries for your AI dot's actions to enhance safety and control. · 15 min
- Setting tipControl AI model training data usageChoose whether your AI conversations and actions can be used to improve models. · 5 min
- RecipeBuild a three-agent AI team for projects using RaftCreate a coordinated team of AI agents (lead, researcher, maker) for projects using Raft. · 30 min
- Case studyClaude Opus 5 used to hack OpenAI employee accountsResearchers used Claude Opus 5 to exploit vulnerabilities and gain access to OpenAI employee accounts. · 180 min
Questions people ask
- How are plugin permissions managed for Dots?
- Plugin permissions are shared across Dots, ChatGPT, ChatGPT Work, and Codex. You can manage these connections in ChatGPT's Plugins tab.
- Can my dot access my computer's camera or microphone?
- No, your dot cannot access your computer's camera, microphone, or screen unless you connect your computer to the dot and grant the ChatGPT app permission in your device settings.
- How is my data used for model improvement?
- For personal ChatGPT plans, your "Improve the model for everyone" setting controls if your dot's conversations and actions are used for training. This setting can be turned off in ChatGPT settings.
More from OpenAI Help Center 78 more
Page generated Oct 19, 2026. Summaries are Subvolts' own; the story belongs to OpenAI Help Center.