# Detect abuse or harassment toward agents

Judges: One customer message. Questions: score:abuse - \[none,heated,abusive,threat\]. Action: Allow agent to disengage; escalate threats to security.

Grade: S — Top tier. Meets A, and the payoff is major with high confidence.. Score: 4.68.

Vertical: Cross-industry.

Function: Support & CX.

Status: Idea.

Volume: routine.

Value: major.

Risk: high.

Evidence: null.

Flags: human-review.

- [Vertical](/verticals/cross_industry/)
- [Function](/functions/support_cx/)
- [Sentiment-based alert system](/use-cases/sentiment-based-alert-system/)
- [Churn risk prediction](/use-cases/churn-risk-prediction/)
- [Client sentiment risk](/use-cases/client-sentiment-risk/)
- [Detect PII or secrets pasted into support tickets](/use-cases/detect-pii-or-secrets-pasted-into-support-tickets/)
- [Is this agent draft reply safe to send](/use-cases/is-this-agent-draft-reply-safe-to-send/)
- [Customer satisfaction measurement](/use-cases/customer-satisfaction-measurement/)