RedactSafe: Ensuring PII Security in Legal Audits

Currently, the management of PII, particularly when sharing and accessing documents on other platforms in this age of digital transformation, is important. In this paper, an integrated framework that attempts to detect and redact PII will be proposed using advanced NLP techniques, namely NER and OCR. These methods are synergistic in order to detect the sensitive data in the structure and unstructured documents. Our method brings together the combination of the conventional rule-based methods and the modern machine learning algorithms in order to dramatically improve the redaction process for both accuracy and flexibility. Regarding enhancing the security and integrity of the data, we apply blockchain, which applies the on-chain storage and verification processes with the strong feature of being immutable and traceable towards redacted documents. We thus establish an access control scheme as hybrid, combining ABAC and RBAC so that access rights of document access get maintained at a suitably secure level and abide to the standards of extant regulations. Experimentation shows results that this framework can function efficiently for data confidentiality preservation without disturbing the functionality in document sharing, which implies an integral approach for sensitive privacy environments. LATEX.