What OpenAI says ChatGPT Enterprise keeps
Every quote in this section comes from an OpenAI page, read on November 13, 2026.
- Retention. "Your workspace admins control how long your data is retained. Any deleted conversations are removed from our systems within 30 days, unless we are legally required to retain them." OpenAI adds that "retention enables features like conversation history, and shorter retention periods may compromise product experience" (OpenAI enterprise privacy).
- Training. "We don't train our models on your organization's data by default." (OpenAI business data)
- Who can read it inside your company. "end users can view their own conversations", and "workspace admins can access an audit log of conversations and GPTs through the Enterprise Compliance API".
- Who can read it at OpenAI. "Authorized OpenAI employees will only ever access your conversations for the purposes of resolving incidents, recovering end user conversations with your explicit permission, or where required by applicable law."
- Security. "OpenAI encrypts all data at rest (AES-256) and in transit".
In short: Enterprise keeps your data for as long as your admins set, encrypted, without training on it by default.
How that compares with zero data retention
Zero data retention means the provider keeps no copy after it answers. OpenAI does offer that, but for its API, the interface developers build on. Its API documentation says the Zero Data Retention control is "subject to prior approval by OpenAI and acceptance of additional requirements" (OpenAI API data controls).
| ChatGPT Enterprise | OpenAI API with zero data retention | |
|---|---|---|
| Who uses it | Staff, in the ChatGPT app | Developers, inside their own software |
| Conversations stored | Yes, for the period your admins set | Customer content excluded from abuse monitoring logs, and the store parameter "will always be treated as false" on covered endpoints |
| How you get it | Through OpenAI sales | "subject to prior approval by OpenAI" |
| Exceptions | Deleted chats can take up to 30 days to leave, or longer if "legally required" | Endpoints marked not eligible "may still store application state" |
The rest of this site explains how zero data retention works and why no training is a different promise.
Was ChatGPT Enterprise caught up in the New York Times order?
No. When a court ordered OpenAI in 2025 to preserve user data in the New York Times copyright case, OpenAI said: "This does not impact ChatGPT Enterprise or ChatGPT Edu customers." It also said API customers on zero data retention endpoints were not affected. OpenAI says its obligations under that order "ended on September 26, 2025" (OpenAI).
That does not mean stored Enterprise data can never be reached. Any data a provider holds can, in principle, be sought through legal process. It means that specific order did not cover Enterprise.
What about the other enterprise chat plans?
The pattern is the same across the big providers: no training by default, storage with admin-set retention.
| Plan | Stored by the provider? | Training |
|---|---|---|
| Claude Enterprise | Yes; "The minimum retention period is 30 days", and "by default, projects are retained indefinitely". Anthropic says the Team and Enterprise interfaces "are not ZDR-eligible" (Anthropic) | "No model training on your content by default" (Claude pricing) |
| Microsoft 365 Copilot | Yes; "The stored data includes the user's prompt and Copilot's response" | "Prompts, responses, and data accessed through Microsoft Graph aren't used to train foundation LLMs" |
| Gemini in Google Workspace | Yes; "90 days to indefinite, as determined by admins" | Not used for model training "outside your domain without permission" |
| Faradex | Files and conversations are held in memory and are gone when the Workspace is reset | Currently Opus via AWS Bedrock, more information is available on our website. AWS does not store prompts or responses, and model providers have no access to them |
The full list of tools that offer zero data retention, by kind of buyer, is on Best Zero Data Retention AI.
Should you worry that Enterprise keeps your data?
That depends on what you put in. For most business work, admin-set retention with no training is a reasonable arrangement, and it gives you conversation history. For confidential client files, ask three questions: how short can our retention be set, who at the vendor can read stored chats, and would we rather the tool kept nothing at all? The FAQ covers the follow-ups people ask most, and Is ChatGPT private? covers the personal and Business plans.