Abstract
Data redaction techniques, such as manual review or simple pattern-matching, may present challenges related to speed, potential for errors, and contextual understanding for identifying sensitive information. The described system and method may be used for a context-aware redaction process that uses a generative artificial intelligence (AI) model. Source text can be encapsulated within a detailed instructional prompt that guides the AI model's function toward identifying and substituting sensitive data. This process can be managed by an orchestration layer that is configured to interact with an AI inference engine, for example, under a zero-retention policy. By replacing identified data with semantic placeholders, the system can generate a modified version of a document. This approach may help redaction outcomes while maintaining aspects of the readability and utility of the text for subsequent analysis or sharing.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
N/A, "System for Context-Aware Data Redaction Using Instruction-Constrained Generative AI", Technical Disclosure Commons, (July 01, 2026)
https://www.tdcommons.org/dpubs_series/10777