Win The AI Wars To Enhance Security And Decrease Cyber Risk

With all the overwrought hype with ChatGPT and AI…much of it earned…you could be forgiven for thinking that only the bad actors are going to be using these advanced technologies and the rest of us are at their mercy.

But this is not an asymmetric battle where the bad actors use AI and the rest of us are struggling using our pencils and abacuses to catch up. It is the good side that invented and is accelerating AI. It is the good scientists that made ChatGPT and all of its competitors. Computer security defenders have been using machine language learning and AI models since the beginning.

At KnowBe4, we have been using and offering machine language- (ML-) and AI-driven technologies for over half a decade. Our customers have been able to use our AI-enabled tools to enhance their security and decrease risk ever since.

AIDA

Our Artificial Intelligence Driven Agent, or AIDA, allows our customers to take advantage of AI-driven training. It allows users to have a more individualized learning experience that adapts to their specific roles, current level of knowledge, phishing and training performance history, and risk factors. KnowBe4’s AI analyzes the unique data and attributes of each of your users, compares against the historical behavioral data of tens of millions of KnowBe4 phishing test users, and identifies and ranks phishing templates that match the skill level of each of your users. Then, the algorithm will select a challenging phishing security test from your admin-approved templates and adapt the degree of difficulty appropriately depending on the unique profile of each user.

And our data shows that, in general, phishing templates chosen by AIDA have a higher rate of receiver clicking than phishing campaigns chosen by human users. But that is good. It is using the power of AI to train your users better, so that they are more able to recognize real phishing attempts, whether generated through AI or a human adversary.

PhishML

Our flagship product, PhishER, uses a feature known as PhishML to analyze reported suspected phishing emails for automatic filtering and prioritization. PhishML helps your InfoSec and Security Operations team cut through the noise and respond to the most dangerous threats more quickly. No suspected email is reported to PhishER until it has already made it through every other defense in your environment. These are the highest risk emails in your environment. PhishML works overtime to help reduce your risk and resource expenditures.

And of course, KnowBe4 is continuing to work overtime to deliver even more products and features using ever more sophisticated models and algorithms.

Our ML- and AI-enabled features have proven to be among the first and best defenses our customers can use. So, do not be scared of ML and AI. Yes, it will certainly be used by bad people to attempt to do bad things. We expect an uptick in both sheer phishing numbers and an increase in sophisticated phishing due to ChatGPT-like malware tools. But the good actors are aggressively fighting back and using better tools to help people fight back and feel safer.

KnowBe4, like other defenders, will be using the best of what ML and AI has to offer along with augmenting with human-based intelligence. Nothing beats the human mind for both creating better defenses and defeating malicious attacks. The answer to AI-based attacks is AI-based defenses along with a well-trained human firewall. AI + humans beat AI alone every time. Come and be a part of that human firewall!