AI Firewall
An AI Firewall is a software layer that detects and redacts sensitive data — PII, PHI, trade secrets, or custom entity types — from a prompt before it leaves your network boundary and reaches any AI model.
An AI Firewall sits between the user (or application) and the AI model provider. It inspects every outbound prompt at the moment of send, identifies values that match sensitive-data patterns, and either blocks the prompt, warns the user, or substitutes the values with reversible tokens.
Backplain's AI Firewall is a patent-pending stateful substitution-and-reconstitution layer. Sensitive entities are replaced with tokens before the prompt leaves the boundary; the model never sees the protected values. On the return path, the response is reconstituted with the original values for the authorized user. Every substitution is logged at the prompt level.
An AI Firewall is distinct from a DLP product (which scans files at rest or in transit) and from provider-side content filtering (which inspects prompts after they have already left your network). It is the only control that prevents sensitive values from ever being transmitted to an external model.
Prompt-time PII redaction is the removal or substitution of personally identifiable information from an AI prompt before it leaves the user's network, so the model and its provider never receive the original values.
Substitution and reconstitution is the technique of replacing sensitive values in a prompt with reversible tokens before transmission, then restoring the original values in the model's response on the return path.
Shadow AI is the unauthorized use of consumer AI tools — typically ChatGPT, Claude, or Gemini on personal accounts — by employees handling work that includes sensitive company or customer data.