detected

Indicates whether custom word content that breaches the guardrail configuration is detected.