BETTER_AI_MEETUP on Can We Outsmart AI? : Cybersecurity in the Age of AI Agents

Join Better_AI_Meetup, where AI designers and builders come together to explore how we can create a better world with AI.

How can useful AI automation become a security risk? And how can a deepfake or recording be used to bypass an online identity check? The 21st Better_AI_Meetup will explore today’s cyber threats and practical ways to defend against them.

Filip Mazán from ESET will share real examples of security risks linked to AI agents and AI-assisted development. He will explain how malicious instructions or excessive permissions can turn useful automation into a liability, and discuss ways to protect sensitive data and monitor agent behavior.

Filip Štiglic from Innovatrics will show how attackers can use deepfakes or genuine recordings to impersonate someone during an online identity check. He will explain how injection attack detection helps verify that images were captured live by a device’s camera, and what to look for when evaluating these defenses.

5:00 pm_Introduction

5:05 pm_Filip Mazán, Director of AI Security Labs at ESET

5:30 pm_Filip Štiglic, Enterprise Architect in Platform Engineering at Innovatrics

5:55 pm_Q&A

6:30 pm_Networking

Both talks and the Q&A will be in English. Find out more and register for Better_AI_Meetup here.