Machine learning and cybersecurity

Agenda:
18:00-18:30: Networking
18:30-18:40: Presenting Kavadev
18:40-19:10: Open Source Models and their Risks, by Natan Katz
19:10-19:40: AI-driven cyber attack simulation, by Teddy Lazebnik
19:40-19:50: break
19:50-20:20: CISO and Data Science, by Salah Masalha

Kavadev Company
With over 20 years of offshore development expertise, Kavadev has positioned Nepal as a premier hub for software solutions. Leveraging a deep understanding of Nepali culture and a robust network of skilled developers, the company ensures top-tier coding standards and reliable project delivery. Kavadev offers cost-effective, enterprise-grade solutions, providing scalable teams that cater to startups and large enterprises alike.

Open Source Models and their Risks
The AI trend in which we experienced was manifested mainly in the SAAS solutions such as OpenAI. The cyber issues that these solutions raised focused on data leakage and privacy. However, in-house and open-source models are still in the neighborhood. Gartner even predicts a massive increase in their commercial usage. This claim is validated when we observe the growth rate of Huggingface and the appearance of Github Models. These models "offer" a whole set of cyber risks that are related not only to the data but to the models themselves. In this lecture, we discuss some of these risks.
Natan Katz is a seasoned expert with over 20 years of experience in Machine Learning and AI research. His work spans various fields, including biometrics, capital market text analysis, and speech processing. In the past three years, Natan has shifted his focus to cybersecurity, working as a researcher and mentor on critical issues such as phishing, IoT blockchain vulnerabilities (specifically in Ethereum), and firewall technologies. Recently, he has embarked on a new venture, leading a startup dedicated to enhancing security for open models.

AI-driven cyber attack simulation
In this talk, we discuss an AI-driven cyber attack simulation that uses static code analysis, graph-based flow models, and a security-games-based decision-making approach to simulate complex and multi-level cyber attacks. We will formalize a mathematical framework for an in silico simulator of cyber-attack simulations upon a single device and a network. Finally, we describe a developing framework and use cases for such a simulator in the development of novel blue and red team strategies.
Teddy is a serial entrepreneur and AI expert. He has a Ph.D. in biomathematics with 60+ academic publications in top-tier journals alongside 85+ successfully delivered algorithmic and AI projects for startups and corporations. Teddy brings over a decade of experience in translating AI into tangible business results. Creating innovative AI solutions for the most complex challenges and directing their business development allowed Teddy to serially provide clients with cutting-edge technology that carefully aligns with their business needs, generating positive ROI quickly and efficiently. Teddy is also an active researcher focusing on personalized clinical treatment design for autoimmune diseases and cancer while also collaborating with a large network of international professionals to bring AI solutions to as many areas as possible.

CISO and Data Science
The data revolution affects many aspects of our lives. On the one hand, it offers incredible services and capabilities. However, it also presents significant challenges. One of the most notable is the transformation occurring in the cybersecurity landscape. Consequently, data science plays a crucial role in modern cybersecurity. In this lecture, I will explore how the CISO and the data scientist must collaborate effectively. I'll approach this topic from the CISO’s perspective.

With 30 years of experience in IT and cybersecurity, Salah is a seasoned Cyber Security Specialist with expertise in leading cybersecurity engineering, R&D, forensic investigations, and secure network planning. He has extensive knowledge in information security, ethical hacking, communication networks, and operating systems, coupled with hands-on experience in IT development and consulting for industrial companies and startups. Salah is skilled in managing high-availability environments, policy setting, and collaborating with global Tier 1 customers. Known for his technical acumen, strong communication, and leadership abilities, he excels in driving cybersecurity initiatives and leading teams effective