At Black Hat USA 2018, IBM unveiled its new AI malware designed to “slip past top-tier defensive measures”.
DeepLocker is “a new breed of highly targeted and evasive attack tool powered by AI”.
The technology company designed the tool to get a better understanding of how existing AI models can be combined with malware techniques to create a “state-of-the-art attack”.
Could DeepLocker outperform a criminal hacker?
The machine learning tool buries itself in other applications until it identifies a victim. When a vulnerability is found the tool ‘unlocks’ it via visual, audio, geolocation and system-level features.
Nevertheless, “we are still far from AI/ML hacking technologies that can outperform the brain of a criminal hacker”, according to High-Tech Bridge, CEO, Ilia Kolochenko.
“It will not invent any substantially new hacking techniques or something beyond a new vector of exploitation or attack. This is because it can be reliably mitigated by existing defence technologies,” he added.
IBM also noted at the Black Hat conference that this method can make it extremely challenging to reverse engineer the benign carrier software and recover mission-critical secrets.
Written by Leah Alger