How Visual Search is Used in Anti-Phishing

Visual Search in Phishing Protection – an effective combination

Visual search is a powerful piece of computer vision technology that can enhance a number of applications, most notably online retail, but now also phishing protection software. In fact, Visual Search may be one of the most integral pieces of technology in ensuring the efficacy of anti-phishing software. 

Evaluating the Need for Visual Phishing Protection

There are many reasons why cyber security companies should incorporate the detection of visual threats into their tech stack. The key one is that it allows the platform to look at an email or web page as the ‘victim’ would see it rather than analyzing the programmatic elements, such as email addresses/URLs, page code, etc.

This is enabled by Computer Vision, which includes a number of technologies, such as logo detection, text detection, object detection and, of course, visual search. 

These technologies combine with existing anti-phishing software to detect visual elements that can be considered suspicious. For example, bad actors will hide key elements like a form as an image so as to evade detection. With computer vision, the attempt to evade detection is nullified. 

Be Sure To Also Read ->Phishing Detection Use Case Overview

How Does Visual Search work in Phishing Protection?

Visual search detects visually similar images to reference images stored in a library. In anti-phishing, this can be used to detect known bad images (KBI), stopping the spread of widely spoofed emails and web pages. It is also a great tool for the detection of known good images (KGI), including payment pages, login forms and so on. The computer vision API renders the page or email as a single graphic. Visual search can then look for and flag any graphical elements that match those in the library, either to a good reference image or a bad one. This establishes an early warning system for any suspicious aspects to an email or page. 

In addition to all this, Image Matching controls can be applied to the technology to ensure that the system looks for only exact images or close matches, depending on your parameters. Reducing the percentage match, the system can highlight variants and new adaptations of known good and bad images.  

Given the critical role that anti-phishing software plays, it’s absolutely essential that the data delivered is as accurate as possible. It’s important to work with a computer vision provider that can provide a precision and recall rate well above 90%. 

The full protection suite 

A combination of text detection, object detection, logo detection, and visual search added to an existing anti-phishing software provides an extra layer of scrutiny and protection, and adds critical signals, allowing for even more accurate decision-making. 

December of 2021 saw the highest number of phishing attacks in almost 20 years prior. Things move so fast in the world of cybercrime that cyber security companies need to take advantage of every technology that can help tip the balance in their favor. 

<h3>Want To Talk??? 

VISUA works with leading cybersecurity companies to empower their technology with Visual-AI. If this is something you are interested in, we’re ready to talk! Fill in the form below, and in the meantime, you can watch our video, in which our CTO explains how Computer Vision and Cybersecurity work together.