Today, we’re excited to announce that Varonis has extended its industry-leading Data Security Platform to OpenAI’s ChatGPT Enterprise. Varonis helps the 3+ million paying users of ChatGPT Enterprise keep their data safe against ongoing risks of misuse, insider threats, and unintended data exposure.
Securing data in ChatGPT Enterprise
ChatGPT takes a different approach from other AI chatbots in how it accesses data. Instead of scanning all corporate data stores, like Microsoft Copilot does, ChatGPT Enterprise uses a ‘connected apps’ approach that only allows access to Google Drive, Microsoft OneDrive and SharePoint files. This integration will honor permissions and allow the user to manually select files from these three resources during the upload action.
However, the challenge remains that users can still download data from any data store and upload content to ChatGPT without the security team knowing the origin or sensitivity of the uploaded data.
In this blog, we’ll discuss how Varonis delivers data security for ChatGPT Enterprise, helping customers avoid privacy and compliance risks by:
- Automatically discovering and classifying sensitive data in sessions
- Monitoring prompts and responses
- Detecting malicious behavior or threats
Automated data classification
Varonis continuously discovers and classifies sensitive data provided to ChatGPT in prompts and the subsequent results with industry-leading accuracy and scale.
Security and risk teams will actively see when users copy and paste PII and other sensitive details directly into prompts and take action. The platform also detects when sensitive data is selected from SharePoint or OneDrive to be processed in a session.
Organizations gain full visibility into how sensitive customer or employee data is being used and accessed, including detailed chat content and intelligence about the user. Varonis provides the context to know whether the information fed into ChatGPT Enterprise is regulated or being used outside the scope of corporate policy.
This up-to-date and accurate picture enables security teams to proactively prevent data breaches and privacy violations.
A single console to visualize risks across AI applications and take action
A single console to visualize risks across AI applications and take action
Continuous session monitoring
Varonis monitors all ChatGPT Enterprise prompts and responses to ensure that sensitive data isn’t uploaded or shared.
Security teams can’t wait for a manually initiated scan or report. Varonis offers a live and comprehensive view of AI activity for forensics analysis and regulatory demands. This feed of AI events is easily searchable and contains a complete history of uploaded files used in conversations.
High-fidelity classification of prompts and responses to detect sensitive data or risky usage
High-fidelity classification of prompts and responses to detect sensitive data or risky usage
Behavior-based threat detection
Built-in threat detection policies automatically alert security teams to potential threats without any configuration required.
New threat detection policies for ChatGPT Enterprise allow security teams to detect abnormal usage and risky changes, including:
- Sensitive file interactions using ChatGPT on OneDrive
- Large file uploads to ChatGPT
- New owner or admin assigned to ChatGPT
Touchless and automated alerting on high-priority AI risks
Touchless and automated alerting on high-priority AI risks
With advanced behavioral analytics, security teams can identify early signs of a breach before it occurs. Varonis provides enriched intelligence on AI activity, enabling them to act on threats without wading through hours of audit logs and fine-tuning an investigation tool.
ChatGPT Enterprise alerts can also stream directly to the Varonis Managed Data Detection and Response (MDDR) team for full investigation and response within an industry-best SLA.
Always-on AI risk defense
Data security is the foundation of AI security. With Varonis for ChatGPT Enterprise, customers get both preventative controls to protect data and active detection of AI risks.
To secure data in the AI era, organizations need current insights along with other data threat signals in a comprehensive platform. Siloed approaches only provide partial visibility by examining a single data source or AI application and offer limited intelligence based on the most recent scan or report.
The Varonis platform provides a centralized solution to safeguard data and AI applications continuously. With the introduction of Varonis for ChatGPT Enterprise, organizations can safely use the AI assistant with always-on defense for every prompt, response, and interaction.
Ready to stop shadow AI and improve your security posture?
Varonis for ChatGPT Enterprise will be available to customers in private preview.
Customers can request a free Varonis Data Risk Assessment to evaluate their AI readiness or request a demo today.
-1.png)
