Runtime Security for AI
Sensitive Data Masking
Sensitive Data Masking
Automatically detect and redact PII, credentials, and financial information in LLM prompts and responses before it reaches the wrong hands.
Automatically detect and redact PII, credentials, and financial information in LLM prompts and responses before it reaches the wrong hands.
The DLP engine scans every interaction for structured and unstructured data, using an extensive library of global identifiers.
Apply masking logic based on your organization’s needs, from full redaction to reversible obfuscation.
Immediately reject prompts or responses containing sensitive data to stop leaks before they happen..
Redact PII and secrets from prompts or completions without interrupting the user experience.
Temporarily mask sensitive data with secure reidentification options for audit or debugging workflows.
Trigger notifications or log events when sensitive data is detected — without blocking the interaction.
Mitigate risks before they escalate through Runtime Security