Padlocked

What constitutes as an AI-Powered Cyberattack?
AI-powered cyberattacks utilise algorithms created via Artificial Intelligence (AI) or Machine Learning (ML) to automate or enhance phases of cyberattacks. Identifying vulnerabilities, advancing attack pathing, establishing system backdoors, and manipulation or exfiltration of data are attacks commonly enhanced with these algorithms.

Types of AI-Powered Cyberattacks
AI is used to adjust, improve or even create ways to conduct cyberattacks, which also include the following below:

• Social Engineering: Algorithms will assist with the researching and executing processes of social engineering attacks in attempts to manipulate human behaviour to fulfil purposes like obtaining sensitive data or granting access to systems. AI will develop the likes of targets to identify and assist with getting into the system, write personalised messages or create multimedia assets like recordings or video footage to gather attention.
• Phishing: AI helps to create highly personalised, realistic emails and outreach that make it easier to fall victim to. They automate real-time communication between phishing attacks as well, like with chatbots having text patterns indistinguishable from humans, posing as customer support for example to obtain sensitive information like account credentials or accessing systems/devices they want.
• Ransomware: With AI, ransomware attacks are optimising performance of the malware to improve it, or help automate its attack path. One example includes AI modifying ransomware files, making them more difficult to detect with cybersecurity tools.
• Deepfakes: These are AI-generated videos meant to deceive others. Commonly used for entertainment purposes via memes on the Internet, they can be used for cyberattacks. Tools with deepfakes can mimic people’s voices and/or faces to help manipulate others into believing what is in front of them.
• Malicious GPTs: Generative Pre-Trained Transformers (more commonly known as GPTs) are AI models that produce intelligent text in response to user-generated prompts. Malicious GPTs intentionally feed users deliberately misinformed outputs, such as supporting attack materials like fraudulent emails to advance attacks.

Mitigating AI-Powered Cyberattacks
Of the many factors to consider with AI-powered cyberattacks, these are the main ways to mitigate them:

• Security Assessments: Deploying and constantly conducting on a strong defensive platform will allow for continuous monitoring, intrusion detection and endpoint protection, whilst also providing real-time analysis of inputs, outputs and abnormal user activity.
• Incident Response Plans: This will outline procedures to follow in the event of a cyberattack. It follows 4 key areas: Preparation (plans to help with preventing and responding to security events), Detection (confirming its occurrence, type and severity), Containment (restricting system use to limit attack’s impact), and Recovery (updating security protocols to safeguard from future threats).
• Employee Awareness Training: Like with social engineering attacks, making employees aware of each of these attacks via a security training course will help mitigate the risk of a company falling victim by educating them of the dangers.