.svg)
ChatGPT Connector for eDiscovery & AI Data Collection
Collect ChatGPT prompts, responses, attachments, and conversation metadata for eDiscovery, compliance, and investigations without disrupting employee workflows. The Onna + ChatGPT Connector provides defensible, enterprise-grade access to AI-generated content across your OpenAI environment.
Why Connect ChatGPT to Your eDiscovery Platform
ChatGPT is rapidly becoming embedded across enterprise workflows for research, drafting, analysis, coding, decision support, and internal productivity. As adoption scales, organizations face a new category of discoverable data: AI-generated interactions.
However, collecting ChatGPT data introduces unique governance and compliance challenges:
AI conversations are not stored like traditional documents
Users interact through persistent and temporary chat sessions
Conversations may contain sensitive business or regulated data
Ephemeral chats are subject to limited retention windows
Data collection depends on OpenAI Compliance API access
Enterprise workspaces may span multiple business units or environments
The Onna + ChatGPT Connector solves this by leveraging the OpenAI Compliance Export API to collect ChatGPT conversations in a structured, defensible format. Organizations can preserve, search, review, and produce AI-generated content alongside traditional enterprise collaboration data.
This enables legal, compliance, and security teams to operationalize AI governance across the enterprise.
ChatGPT Connector Capabilities
The Onna + ChatGPT Connector is purpose-built for enterprise AI governance and large-scale AI data collection.
Key capabilities include:
Secure authorized connection using OpenAI API credentials
Workspace-based ChatGPT Enterprise collection
Custodian-based conversation collection
Audit logs for all collection activity
Support for persistent and ephemeral conversations
Structured ingestion of prompts and responses
Incremental sync using date-based checkpoints
One-time sync, auto-sync, and archive sync modes
Attachment collection for uploaded and generated files
Metadata capture including GPT model versions
Searchable standalone conversation threads
Identity mapping for custodian attribution
Support for multiple ChatGPT Enterprise workspaces
These capabilities help organizations create scalable, defensible AI governance workflows across legal hold, investigations, compliance, and regulatory response initiatives.
What Data Can Be Collected from ChatGPT
The connector captures structured AI interaction data from ChatGPT Enterprise and ChatGPT Edu environments through the OpenAI Compliance API.
Conversations
Full ChatGPT conversation threads
Chronological prompt and response sequences
Persistent and ephemeral chat sessions
Standalone reviewable conversation records
Edited prompt history and revisions
Messages (Per Turn)
User prompts submitted to ChatGPT
AI-generated responses
Message-level timestamps
Prompt edit tracking
Prompt-response sequencing
Participants
Workspace users (custodians)
User IDs and email attribution
ChatGPT system responses
Custodian-linked conversation grouping
Attachments
User-uploaded files
ChatGPT-generated files
PDFs, DOCX, CSVs, images, and other supported content
Content Structure
Conversation threads preserved in chronological order
Clearly distinguishable human and AI interactions
Independent searchable conversation records
Structured ingestion for downstream review workflows
This structured model enables investigators and legal teams to reconstruct AI-assisted decision-making and collaboration workflows with full conversational context.
ChatGPT Metadata Collected
In addition to standard metadata, the connector captures AI-specific metadata attributes including:
Conversation IDs and thread identifiers
Workspace and custodian attribution
User IDs, names, and email addresses
Conversation titles
Prompt and response timestamps
GPT model version information
Edit history metadata
Source attributions from web-enabled responses
File attachment metadata
Sync and export timestamps
Workspace source attribution
Examples of captured model metadata include GPT-4o and GPT-4 Turbo usage within conversations.
This metadata enables advanced filtering, timeline reconstruction, custodian analysis, and model-level reporting across enterprise AI usage.
Onna + ChatGPT Connector Requirements
To connect ChatGPT Enterprise or ChatGPT Edu to the Onna platform, the following requirements must be met:
Active ChatGPT Enterprise or ChatGPT Edu subscription
OpenAI Compliance API enabled for the workspace
OpenAI-issued API key for the organization
ChatGPT Workspace ID
Workspace owner or admin permissions in OpenAI
Authorized connection configured in Onna
The ChatGPT connector is not compatible with ChatGPT Free, Plus, or Team plans because the Compliance API is only available to Enterprise and Edu customers.
How ChatGPT Data Collection Works
The connector uses the OpenAI Compliance Export API to collect ChatGPT conversation data on a per-custodian basis.
Configure an Authorized Connection
Administrators configure an authorized connection within Onna using:
- OpenAI API Key
- ChatGPT Workspace ID
- Enterprise source configuration
Each ChatGPT Enterprise workspace requires its own dedicated connection
Resolve Custodians
Onna maps workspace users to OpenAI user identities and resolves:
- User names
- Email addresses
- Custodian relationships
- Workspace attribution
Request Data via the OpenAI Compliance API
Onna submits collection requests to the Compliance Export API using configured date filters and custodian criteria.
The API returns:
- Persistent conversations
- Temporary and ephemeral chat sessions
- Updated conversations within the selected timeframe
Because the API supports a “greater than or equal to” date operator, collections retrieve all qualifying content from the selected start date forward.
Download and Normalize Data
Onna downloads exported conversation data and structures it into searchable records.
This includes:
- Prompt-response threading
- Metadata normalization
- Attachment ingestion
- Conversation indexing
Search and Review
Collected ChatGPT conversations become immediately searchable within the Onna platform and can be:
- Reviewed
- Tagged
- Exported
- Included in legal hold workflows
- Used in investigations and compliance reviews
ChatGPT Data Collection Options
The connector supports flexible sync configurations for different legal and compliance workflows.
One-Time Sync
Targeted collection for investigations, litigation matters, or custodian-specific reviews.
Auto Sync
Continuously captures newly available ChatGPT conversations and updates.
Auto Sync + Archive
Maintains a defensible archive of AI-generated interactions for long-term governance and compliance programs.
Common ChatGPT eDiscovery Use Cases
AI Governance & Oversight
Monitor and govern employee use of generative AI across the organization.
Litigation Response
Collect AI-generated content relevant to legal matters and case strategy.
Internal Investigations
Review prompts, responses, and attachments tied to misconduct, insider risk, or policy violations.
Regulatory Compliance
Support emerging AI governance requirements and defensible AI data retention policies.
Data Security & Risk
Identify sensitive information exposure within prompts, generated outputs, or uploaded files.
IP Protection
Preserve records of AI-assisted drafting, coding, research, and content generation activities.
ChatGPT Connector Considerations & Limitations
Enterprise Subscription Required
The connector only supports ChatGPT Enterprise and ChatGPT Edu environments. Free, Plus, and Team plans are not supported.
Compliance API Availability
Organizations must confirm that the OpenAI Compliance API is enabled for their workspace before configuring the connector.
30-Day Retention Window
Ephemeral and deleted conversations are only available within OpenAI’s 30-day retention window unless otherwise retained for legal or regulatory reasons.
Date Filtering Constraints
The Compliance API currently supports start-date filtering only. Hard end-date filtering is not available at the API level.
Multiple Workspace Support
Organizations with multiple ChatGPT Enterprise workspaces must configure separate authorized connections for each environment.
No Real-Time Collection
Collections reflect data available through the Compliance API at the time of sync and do not provide real-time ingestion.
%201.webp){kind=logo}
%201.webp)
Onna + ChatGPT Connector FAQs
Yes. The connector uses the OpenAI Compliance Export API to retrieve ChatGPT Enterprise and ChatGPT Edu conversation data.
The connector collects prompts, responses, attachments, timestamps, metadata, model information, edit history, and conversation threads as structured records.
Yes. Ephemeral chats are collected if they fall within OpenAI’s 30-day retention window.
Yes. The connector supports custodian-based collection workflows.
Yes. User-uploaded files and ChatGPT-generated files are collected alongside conversation content.
Yes. Each workspace requires its own authorized connection and configuration.
No. Collection timing depends on data availability through the OpenAI Compliance API.
Yes. Conversations involving internally developed custom GPTs are collected through the Compliance API.
Start Collecting ChatGPT Data
Connect ChatGPT Enterprise in minutes and begin collecting AI-generated conversations, prompts, responses, and attachments across your organization for eDiscovery, compliance, investigations, and AI governance.