# 1 : Safety vs. Security for AI of Autonomous Cars?

After extensive research and development, the automobile industry is on the cusp of a new era in computing with the emergence of semi and fully autonomous cars, which offer automatic alternatives to traditional manual driving.

However, as we approach this transformational shift, critical questions arise concerning the risks associated with the AI that powers these autonomous vehicles. How can we ensure the security of the AI systems controlling these cars? And is it possible that a seemingly more secure AI could paradoxically lead to more dangerous driving scenarios?

In a compelling talk, join Ben as he delves into the safety and security challenges associated with autonomous vehicles. Ben will shed light on how seemingly innocent objects like digital billboards and emergency vehicles, encountered frequently while driving, can actually jeopardize the safety of both the autonomous car's occupants and other road users. He will present video demonstrations illustrating these risks specifically concerning Tesla vehicles.

Furthermore, Ben will meticulously analyze the existing security mechanisms in place and argue for the necessity of developing new and dedicated approaches to secure the AI driving these autonomous vehicles which can withstand potential threats and guarantee passenger safety.

Lastly, Ben will address a counter-intuitive belief: that enhancing AI security might somehow lead to more dangerous driving conditions. Through thoughtful analysis, Ben will explain how greater AI security can, in fact, contribute to less safe and more dangerous driving. Join us for this enlightening discussion on the safety and security challenges of the autonomous vehicle revolution.

# 2: AI is coming.. Are you ready?

We have officially entered the era of AI, where prominent companies like OpenAI, Google, Tesla, and others have seamlessly integrated AI models into both physical (e.g., autonomous vehicles) and digital systems (e.g., chatbots). These AI-powered systems continuously and extensively interact with end users, presenting exciting opportunities and cutting-edge functionalities.

However, this revolutionary integration of AI also introduces unprecedented risks that demand our attention. What do we truly understand about these new risks posed to end users by AI?

In this talk, Ben delves into the challenges arising from integrating AI into physical and digital systems that heavily interact with end users. Ben conducts an analysis of the new threat models emerging from the interface between AI and end-users. Through compelling videos, he demonstrates how attackers can exploit these systems:

• By implanting malicious banners on websites to target users of an LLM.
• By compromising advertisements displayed on digital billboards to attack drivers and passengers of autonomous vehicles.

During the talk, Ben thoroughly examines the current security mechanisms employed to protect end users from potential threats. Despite their existence, he highlights the imperative need for new and dedicated security mechanisms specifically tailored to safeguard end users against AI-based systems like chatbots, autonomous vehicles, and others. As the talk reaches its conclusion, Ben focuses on discussing the precautions that end-users must adopt while interacting with AI-based systems. By embracing proactive measures, we can ensure the safe integration of AI into systems that extensively interface with end users.

# 3: Making Smart Cities Safer

The 'Smart Cities Revolution' has witnessed numerous cities worldwide embracing cutting-edge advancements such as medicine deliveries by drones, rides conducted by robotic taxis, and the implementation of green technologies to conserve resources and money. Cities like San Francisco, Barcelona, Tel Aviv, and many others have actively participated in this transformation.

In his enlightening talk, Ben Nassi addresses the critical challenges to security and privacy arising from the 'Smart Cities Revolution'. He endeavors to answer a fundamental question: Can society and new urban technology (e.g., drones, autonomous vehicles) truly take their relationship to the next level while ensuring security and privacy?

Ben meticulously analyzes the novel security and privacy risks that accompany the deployment of drones, autonomous vehicles, and green technology in smart cities, elucidating their far-reaching consequences. He emphasizes the need for a paradigm shift in security and privacy practices, transitioning from traditional methods used to safeguard data centers to new security mechanisms tailored to protect the safety and privacy of end users and urban technology in the physical world.

During his talk, Ben thoughtfully explores the case of deploying drones in the physical world, shedding light on the delicate tradeoff between the security and privacy of these aerial devices and that of society as a whole. By delving into these essential aspects, he offers valuable insights into how to navigate the security and privacy challenges posed by the rapid advancement of technology in our urban environments.