Addressable Smart Agents for Adaptive Unauthorized Transaction Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional fraud and abuse detection systems in large computer networks struggle with scalability, require high manual effort, and fail to adapt to evolving threats, often missing unusual behaviors of system administrators and hackers due to rigid modeling and the need for constant retraining with static datasets.

Innovation Solution

Implementing smart agent technology with artificial intelligence and machine learning to analyze and profile the behavior of system administrators, using addressable smart agents to detect deviations from normal behavior patterns and trigger alerts for unauthorized activities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional neural network models are used for fraud detection, then detection capability is provided, but the system requires constant retraining with static datasets and cannot adapt to evolving threats

Engineering Contradiction:

Improveadaptability to evolving threatsVSAvoidtime for periodic model redevelopment

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent transforms static fraud detection models into dynamic systems that continuously learn from incoming transactions. The smart agent creates and updates behavioral profiles in real-time, allowing the system to adapt to evolving fraud patterns without periodic retraining interruptions. This dynamic approach enables the system to evolve alongside new fraud techniques.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements self-service learning where the fraud detection model automatically improves itself through continuous analysis of transaction data. The smart agent autonomously updates behavioral profiles and detection parameters without requiring external retraining interventions, enabling the system to self-adapt to new threats as they emerge.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If neural networks are trained with historical transaction data, then fraud detection accuracy is improved, but the system fails to detect new types of fraud that deviate from historical patterns

Engineering Contradiction:

Improvefraud detection accuracyVSAvoiddetection of new fraud types

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent establishes baseline behavioral profiles for legitimate transactions before fraud occurs. These pre-established profiles serve as reference points that enable the system to detect deviations indicating new fraud types. By having preliminary behavioral understanding, the system can identify novel threats that diverge from normal patterns.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where detection results and new transaction patterns feed back into profile updates. This feedback mechanism allows the system to learn from both confirmed fraud cases and legitimate transaction variations, continuously improving its ability to detect new fraud types while maintaining accuracy.

Inventive Principle:

Principle #23Feedback

3Ease of operation

If system administrators are given super access to all parts of the network, then system management capability is improved, but the risk of unauthorized activity and security damage increases

Engineering Contradiction:

Improvesystem management capabilityVSAvoidsecurity risk from admin abuse

Core Design Contradiction:

Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The patent introduces smart agents as intermediary monitoring layers between system administrators and network resources. These agents continuously observe administrator behaviors and compare them against established profiles, acting as intermediaries that enable full system access while simultaneously detecting and alerting on suspicious activities. This mediator approach preserves management capability while mitigating security risks.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Reliability

If conventional fraud detection systems monitor all transaction activities, then detection coverage is improved, but computational resources and processing time increase significantly

Engineering Contradiction:

Improvedetection coverageVSAvoidcomputational resource consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent applies local quality by creating individualized behavioral profiles for different entities (cardholders, merchants, terminals) rather than using uniform monitoring rules. Each entity has its own profile capturing its specific transaction patterns, allowing the system to focus computational resources on detecting deviations from individual baselines rather than applying blanket monitoring to all transactions equally.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS12386952B2Addressable smart agent data technology to detect unauthorized transaction activity

Publication Date: 2025.08.12 BRIGHTERION INC

AI summary

A computer implemented and electronic process is provided that uses artificial intelligence to detect unauthorized activity by an insider or hacker. Electronic systems that employ artificial intelligence and machine learning to detect unauthorized transaction activity by insiders or hackers for a computer network system are also provided. Hardware required for carrying out the invention typically include a plurality of networked computers. Specialized software and/or firmware is typically needed in connection with the hardware for carrying out the invention.