Rockfort Orion
Employee-facing LLM DLP delivered via a browser extension. Prevent sensitive data from reaching external LLMs (ChatGPT, Claude, etc.), reduce Shadow AI risk, and give InfoSec complete visibility.
Note: Orion is employee-facing and designed to work quietly in the background without disrupting teams.
Agent Prompt
Customer: Jane Doe
SSN: 123-45-6789
Card: 4242 4242 4242 4242, Exp: 10/27, CVV: 123
Issue: Customer is requesting a refund for order #A12345 placed last week.
Action: Draft a polite response, verify details, and propose next steps.Detected: SSNDetected: Credit CardName
ChatGPT
Pending
Awaiting redacted prompt...
Real-time Data Loss Prevention
Detects and blocks sensitive data before it’s sent to external LLMs in the browser.
Shadow AI Risk Reduction
Covers unmanaged usage of GenAI tools to reduce data exposure across employees.
Enterprise Controls
Targeted policies by department, role, domain, or app with enforce or warn modes.
Privacy First
Runs locally in the browser where possible; minimal telemetry, configurable redaction.
How Orion works
Browser-native protection
Orion’s extension monitors text boxes and requests to known LLM domains. Sensitive content is flagged and redacted or blocked before leaving the device.
Policy engine
Define patterns and rules (PII, secrets, customer data, project names) with modes: monitor, warn, block. Scope policies by group, domain, or application.
Visibility and controls
Centralized reporting for InfoSec. Integrates with your identity provider for role-based controls and offers exportable logs for audit.
Visibility for InfoSec
Usage Analytics
Understand which GenAI tools are used, by whom, and for what types of tasks.
Policy Insights
Track blocked/redacted events and measure policy effectiveness over time.