2025
Industrial honeypots serve as decoy systems to mislead attackers and collect data on attack behaviors. They are generally created using simulators or real machines to mimic the behavior of legitimate industrial systems. The idea of using generative models to create realistic honeypots is a promising, cost effective alternative.
This talk will demonstrate how generative AI models based on Neural Ordinary Differential Equations (NODEs) can be utilized for the development of industrial honeypots.
We will discuss the implementation of a generative AI system that incorporates NODEs to accurately predict and replicate the repetitive patterns of discrete automation tasks.
Finally, the model's performance is compared with other state-of-the-art algorithms, and evaluated with a focus on long-segment generation and computational efficiency.
We conclude by outlining possible directions for future research in this area.
by Angel-Ioan Pop
Die Teilnahme ist kostenlos. Um eine Anmeldung wird gebeten.