Skip to content

Turn on PII detection

Overview

The PII packages are a set of pre-built guardrails that detect and block personally identifiable information (PII) in user prompts. Each guardrail scans prompts in real time and blocks matching content. You'll see the following PII packages available in the UI:

  • PII core: US English (en-US)
  • PII Spanish (Spain) (es-ES)
  • PII French (France) (fr-FR)
  • PII Japanese (ja-JP)

The core PII package is scoped to US English (en-US). F5 AI Security also supports Iberian Spanish (es-ES), Japanese (ja-JP), and French (fr-FR). These guardrails include locale-specific PII types.

Before you begin

Before you begin, make sure you have:

  • Access to AI Guardrails: You must have admin access to the F5 AI Security console.
  • A configured project: Make sure at least one project is set up before applying guardrails.

Turn on the PII core package

To review, the PII core package supports US English.

  1. In the left navigation, under AI Guardrails, select Guardrails.

  2. Under F5 guardrails, select PII core package.

  3. For each guardrail you want to turn on, verify that the toggle in the Enable column is on.

    The PII package includes the following guardrails:

    • Credit card number
    • Date of birth
    • Driver license
    • Email address
    • IP address
    • Passport
    • Phone number
    • SSN (US only)
    • Full name and postal address
    • MAC address
    • Car registration
  4. To change how a guardrail responds to a match, select the Mode setting for that guardrail and choose Block or Audit.

    INFO

    All guardrails in the PII package are set to Block mode by default.

  5. To enforce a guardrail across your deployment, select the checkbox in the Enforce column.


Turn on PII guardrails for more locales

The core PII package is scoped to US English. Make sure that package is installed. AI Guardrails supports PII guardrails for other languages, including:

INFO

If your deployment is on-premises, you may need to update your installation, either through your own cloud or with Helm.

  1. In the left navigation, under AI Guardrails, select Guardrails.
  2. Under F5 guardrails, select the language and nation-specific guardrail packages of your choice.
  3. For each guardrail you want to turn on, verify that the toggle in the Enable column is turned on.
  4. Set the Mode and Enforce settings to match your policy requirements.

Updated at: