Neuro-Symbolic API Access Assessment for Partial Data Sharing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large organizations face challenges in efficiently, securely, and uniformly managing information exchange between internal and external computer systems, particularly in open banking scenarios where APIs are vulnerable to unauthorized access and lack robust digital identity validation.
Innovation Solution
A neuro-symbolic artificial intelligence-based assessment system that uses neural networks and symbolic reasoning to identify access patterns, build sub-objects in real time, and perform partial data transfers to ensure secure data sharing, maintaining an abstract enterprise system of record and a separate shareable storage system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional API authentication methods (service ID and token checks) are used, then service call validation is achieved, but digital identity validation is insufficient and vulnerable to token forgery
Solution Approach 1:
The patent introduces an intermediary assessment system that sits between the API caller and the target system. This intermediary performs digital identity validation using neural networks and symbolic reasoning, preventing direct authentication attacks while maintaining system security without requiring complex changes to existing API endpoints.
Solution Approach 2:
The patent replaces traditional mechanical authentication mechanisms (service ID and token checks) with an AI-based assessment system using neural networks and symbolic reasoning. This substitution provides more robust digital identity validation that can detect forged tokens and assess caller intent without relying on easily compromised cryptographic tokens.
2Productivity
If complete data transfer is performed for API calls, then information sharing is efficient, but data security and access control are compromised
Solution Approach 1:
The patent segments the data transfer process into multiple phases based on assessment results. Instead of transferring complete datasets, the system divides data into authorized portions and unauthorized portions, transferring only the authorized segments after neural network assessment of the caller's digital identity and intent.
Solution Approach 2:
The patent implements partial data transfer by providing only the subset of data that the caller is authorized to access based on digital identity assessment. This partial action approach maintains security by withholding unauthorized data while still providing efficient service for authorized requests, avoiding the need to transfer excessive data and then filter it.
3Ease of operation
If open banking APIs are exposed to external interfaces, then user experience is improved, but unauthorized access and fraud risk increase
Solution Approach 1:
The patent applies preliminary anti-action by performing digital identity validation and fraud assessment before allowing any data access through open banking APIs. The neural network assesses the caller's intent and authenticity in advance, preventing unauthorized access and fraud before they can occur, while still maintaining easy access for legitimate users.
Solution Approach 2:
The patent implements feedback mechanisms where the assessment system continuously monitors API calls, evaluates digital identities, and adjusts access decisions based on detected patterns and behaviors. This feedback loop enables the system to maintain ease of operation for legitimate users while dynamically blocking fraudulent activities without requiring users to be aware of the security measures.
Data Source
AI summary
Various aspects of the disclosure relate to dynamically determining user access levels to manage access to enterprise information via application programming interfaces (APIs). A neuro-symbolic AI-based assessment enabled system manages assessments and response to API calls to ensure data security of information shared with external sources via the API. This system identifies and analyze access patterns via neural networks and symbolic reasoning to dynamically manage a rule set to determine access levels and corresponding data sub-objects that are built in real time to be shared with an API response message.