Least-Privilege Permission Management via Predictive Usage Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The complexity of cloud services introduces security vulnerabilities due to broad permissions granted to user accounts, applications, or processes, leading to unintentional and malicious changes and data leaks.

Innovation Solution

A system that predicts the future resource usage of an identity by analyzing its history, the history of similar identities, and peer identities, then determines the bare minimum permissions needed to perform these predicted actions, using a graph-based model with node embeddings and deep learning techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If broad permissions are granted to user accounts, applications, or processes to ensure they can perform all possible tasks, then operational flexibility and productivity are improved, but security vulnerabilities increase leading to unintentional and malicious changes and data leaks

Engineering Contradiction:
Improveoperational flexibilityVSAvoidsecurity vulnerabilities
Core Design Contradiction:
Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The patent implements dynamic permission adjustment by continuously monitoring resource usage patterns and automatically updating permission levels. The system transitions from static broad permissions to dynamic least-privilege permissions that adapt based on observed behavior, resolving the contradiction between operational flexibility and security by making permissions flexible only when necessary.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system performs preliminary analysis of resource usage patterns before making permission decisions. By studying historical usage data and predicting future needs, the system pre-configures appropriate permission levels, avoiding both excessive broad permissions and overly restrictive limitations, thus balancing security with operational flexibility.

Inventive Principle:
Principle #10Preliminary action

2Device complexity

If permission analysis is based only on past resource usage history, then implementation simplicity is improved, but accuracy of predicting future resource usage needs deteriorates

Engineering Contradiction:
Improveimplementation simplicityVSAvoidprediction accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system performs preliminary analysis of resource usage patterns before making permission decisions. By studying historical usage data and predicting future needs, the system pre-configures appropriate permission levels, avoiding both excessive broad permissions and overly restrictive limitations, thus balancing security with operational flexibility.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements dynamic permission adjustment by continuously monitoring resource usage patterns and automatically updating permission levels. The system transitions from static broad permissions to dynamic least-privilege permissions that adapt based on observed behavior, resolving the contradiction between operational flexibility and security by making permissions flexible only when necessary.

Inventive Principle:
Principle #15Dynamics

3Measurement precision

If deep learning models with node embeddings are used to identify similar identities and predict resource usage, then prediction accuracy is improved, but system complexity increases

Engineering Contradiction:
Improveprediction accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system employs self-service mechanisms where the deep learning model automatically trains on organizational data and continuously improves its predictions without manual intervention. The model serves itself by learning from patterns in the data, reducing the need for complex manual configuration and expertise while maintaining high prediction accuracy.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual permission management processes with automated deep learning-based prediction systems. Instead of security administrators manually analyzing usage patterns and making permission decisions, the system uses machine learning models to automatically perform these tasks, substituting mechanical human processes with automated computational ones.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4100857B1Least-privilege resource permission management
Publication Date: 2025.02.26 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP4100857B1 patent drawingFigure 1
  • EP4100857B1 patent drawingFigure 2A~2B
  • EP4100857B1 patent drawingFigure 3

AI summary

The least-privilege permission needed for an identity, such as a user account, application, user group, or process, to access a resource of a tenant of a cloud service is determined from a predicted future resource usage. The predicted future resource usage is based on the resource usage history of an identity, the resource usage history of similar identities and the resource usage history of its peers. Similar identities are determined from node embeddings of a graph that represents the assigned permissions of an identity to a resource and the usage activity at a resource. The permissions needed to perform the predicted future resource usage is compared with the current permission assignments to determine the bare minimum permission that an identity needs for its ongoing and future workflow.