Data masking can be used for Anonymization or Pseudonymization to comply with GDPR, Paper Redaction and Data Preprocessing to feed NLP models.