LLM Semantic Access Control Labeling for Enterprise Documents

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems struggle to accurately set access controls for documents in enterprise networks, leading to data loss due to incorrect implementation by non-technical personnel and poor performance in handling documents targeted towards different audiences.

Innovation Solution

Utilizing a Large Language Model (LLM) for semantic understanding to extract ideas from documents, compare them with existing documents, and apply the access control lists of similar documents to new documents, thereby automating the access control process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If access control is manually set by non-technical personnel, then access control can be configured, but it is frequently done incorrectly leading to data loss

Engineering Contradiction:

Improveaccess control accuracyVSAvoidmanual configuration difficulty

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system enables self-service access control by having the document itself provide access control information through its metadata and content. The access control mechanism automatically extracts and applies access control lists from document metadata without requiring manual configuration by users, thereby eliminating human error while maintaining ease of document sharing.

Inventive Principle:

Principle #25Self-service

2Reliability

If access control is set to permit only relevant users, then security is improved, but documents become hard to share

Engineering Contradiction:

Improveaccess control securityVSAvoiddocument sharing efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system implements feedback mechanisms where the access control mechanism continuously monitors document access patterns and automatically adjusts access control lists based on actual usage. This allows the system to maintain security by blocking unauthorized access while enabling legitimate sharing by learning from document access feedback and dynamically updating permissions.

Inventive Principle:

Principle #23Feedback

3Reliability

If automated access control systems are used, then manual configuration errors are reduced, but performance is poor for documents targeted towards different audiences

Engineering Contradiction:

Improveaccess control consistencyVSAvoidaudience-specific access control

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system applies local quality by extracting and applying access control information specifically from the metadata and content of each individual document rather than using a single centralized access control policy. This allows the access control mechanism to adapt to different document types, audiences, and sensitivity levels, thereby maintaining consistency while improving versatility for audience-specific documents.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260067281A1Access control labeling via LLM semantic understanding

Publication Date: 2026.03.05 CISCO TECHNOLOGY INC

AI summary

In one implementation, a device extracts, using an embedding model, one or more ideas from a particular document. The device determines a measure of similarity between the one or more ideas from the particular document and those of each of a body of existing documents, to identify a set of one or more similar documents. The device generates an access control list for the particular document, based on one or more access control lists associated with the set of one or more similar documents. The device restricts access to the particular document according to the access control list for the particular document.