I recently published an article entitled “Artificial Intelligence: The Cyber Weapon of the Future” in which I highlighted the rapid development of Artificial Intelligences (A.I) technology across industries as well as the opportunity and threat it represents. In the wrong hands, A.I can be used for malicious purposes by breaking through various traditional security measures.
IBM Basic Research Laboratories states cybercriminals are already weaponising A.I and the best way to counter such a threat is to learn how it works. In response to this growing threat IBM developed “DeepLocker”- next generation malware powered by A.I that has the unique ability to mask its malicious intent until it hurts a specific target.
This A.I. powered malware is particularly threatening because it could be hidden in any standard application, like video conferencing application, to avoid detection by traditional antivirus software. As a result it could easily infect millions of systems simultaneously without being detected. According to IBM, DeepLocker “unleashes its malicious action as soon as the AI model identifies the target through indicates like facial recognition, geo-location and voice recognition”.
How it Works?
For demo purposes at Black Hat conference in Vegas, USA, IBM hid the malware in a video conferencing application and trained the A.I model to unleash it in response to facial recognition.
When the malware detected the target using a webcam, it provided the indicator to lock down the victim’s system.
In this case, DeepLocker requires only a photo of you which can be easily extracted from any of you social media channels ( Facebook, LinkedIn..)
Based on their findings, IBM is actively developing countermeasures using behavior-based technology to identify when an application is deviating from the norm and by using cyber-deception techniques to trick A.I powered malware.
Awareness if the first layer of defense. A.I powered threats are coming our way very, very soon and we need to be prepared.