Question
During a security review after a support-data exposure incident, a company wants to analyze ticket text with AI to identify complaint categories. The text includes account numbers, health-related notes, and employee comments copied from internal systems. What should the team do before using the text?