Ransomware vs. AI. Part 2 – Bypassing Ransomware Protection with Reinforcement Learning
Tid: 19 augusti 2020 09:00-09:45
Plats: BTH Campus Gräsvik
The Professional Masters in Information Security project (PROMIS -promisedu.se) presents online inspiration seminars on cutting edge topics and areas.
At the end of the seminars we will shortly also discuss and show our new educational initiative, and courses specifically adapted towards the security area FOR industry/working active people looking to further their knowledge in the area. The courses are full university courses but given in a manner so that you can take them on distance without disturbing your work!
For details on PROMIS: www.promisedu.se
For details in the seminars see below: (no registration required!).
Join in via Zoom link: https://bth.zoom.us/j/66653906023
Or via Zoom Meeting ID: 666 5390 6023
“Ransomware vs. AI. Part 2 – Bypassing Ransomware Protection with Reinforcement Learning”
Reinforcement learning is a well-known concept used in the gaming industry, mostly in real-time strategies, to support computer players with AI. It is also known that AlphaGo, a computer program designed by DeepMind, has been recognized as the best Go and chess player in the world. The algorithm was trained by playing with human players as well as with other instances of itself to improve its play.
So, what if we use RL in Security Testing to discover vulnerabilities in products and weaknesses in defense? To prove that hypothesis, we applied RL to run ransomware simulation that can learn how to bypass anti-ransomware protection. We’ll present our results in the presentation.