Skip to main content

Introducing WonderFence

What is WonderFence?​

WonderFence is a real-time AI guardrails platform that helps organizations monitor, control, and enforce safety and security policies across their AI systems. It sits between your users and your AI applications — inspecting inputs and outputs as they flow through your systems — to detect, block, or redact content that violates your safety and security policies.

Whether you operate customer-facing chatbots, internal AI assistants, or generative AI workflows, WonderFence provides the enforcement layer that ensures your AI systems behave within the boundaries you define.

Key Use Cases​

  • Content moderation — Detect and block toxic, violent, sexual, or otherwise harmful content in AI-generated responses before it reaches users.
  • Policy enforcement — Define and enforce organization-specific content policies across all AI touchpoints, ensuring consistent behavior at scale.
  • Data leakage prevention — Identify and redact PII, credentials, internal system details, and other sensitive information from both user inputs and AI outputs.
  • Jailbreak and prompt-injection defense — Detect adversarial inputs designed to bypass safety instructions and prevent them from reaching your models.
  • Regulatory compliance — Maintain enforceable policy records and audit trails that demonstrate adherence to industry standards and AI-governance frameworks.

Note — Before you continue, make sure that you have set up the data flows between your platform and WonderFence, as described in the Alice Integration and API Guide, which can be accessed by clicking the API DOCUMENTATION link that appears in the bottom-left corner of the page when the main menu is displayed.