Outsmarting AI

One can think of many arguments of why machine learning (ML) is a game changer in the field of cybersecurity. In fact, today, many ML applications already exist to protect our systems from attacks in more intelligent ways than before. However, have you considered that your own ML model can be the direct target of attacks? This talk elaborates on the topic of adversarial attacks in ML; How they work and how you can defend against them. 

Key takeaways:

• Know how your ML model/application works
• Human interaction is crucial for AI-first autonomous systems
• Defence against adversarial attacks should be built into your ML application design

Presentation deck

Outsmarting AI

Click here to download the slide deck. Please note that downloads are only available for event participants and subscribers. You'll need to log in to download it.