Browser-in-the-Browser Attack Protection Using AI Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Browser-in-the-Browser (BiTB) attacks simulate login windows to steal user credentials, exploiting single sign-on (SSO) options and OAuth protocols, posing a significant security risk as existing technologies struggle to verify the legitimacy of websites and detect spoofed domains effectively.
Innovation Solution
A website connection verifier (WCV) module operates in the background, capturing screenshots, analyzing images for pop-ups and network addresses, and verifying authentication requests with authenticator websites to confirm legitimacy, using AI for image recognition and database comparisons to identify spoofed sites.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing technologies are used to verify website legitimacy, then the verification process is simple, but the detection precision of spoofed domains is insufficient
Solution Approach 1:
The verification system is segmented into multiple independent components: screenshot capture module, AI image analysis module, database comparison module, and verification module. Each component performs a specific function, allowing the system to achieve high detection precision through coordinated operation of specialized modules without requiring each individual module to be overly complex
Solution Approach 2:
The patent introduces intermediary elements including AI models as intermediaries between the screenshot and domain verification, and database records as intermediaries for comparing expected versus actual domain information. These intermediaries enhance detection precision by adding layers of analysis without significantly increasing overall system complexity
2Reliability
If screenshot capture and AI analysis are implemented, then the detection capability improves, but the processing time increases
Solution Approach 1:
The system performs preliminary actions by capturing screenshots immediately when authentication prompts appear and pre-processing images through AI analysis before final verification. Expected domain information is pre-stored in databases, enabling rapid comparison and reducing real-time processing requirements, thus maintaining high reliability while minimizing time loss
Solution Approach 2:
The verification process skips unnecessary intermediate steps by directly analyzing authentication prompt screenshots and comparing extracted domain information against expected values in the database. The system rushes through the verification by focusing only on critical security-checking operations rather than comprehensive website analysis, improving speed without sacrificing reliability
Data Source
AI summary
Aspects of the subject disclosure may include, for example, capturing an image of information displayed on a display screen of the device, identifying in the image a pop-up, the pop-up prompting entry of user credentials of a user for an authenticator website, the pop-up generating an authorization request from the device to the authenticator website, identifying in the image a network address for a target website, and communicating with the authenticator website at the network address for verification of the authorization request. Other embodiments are disclosed.