Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the critical landscape of AI security at the edge in this 41-minute conference talk featuring Mounia Kharbouche from STMicroelectronics at EDGE AI Taipei 2025. Delve into the fundamental question of AI model trustworthiness as experts examine why blind trust in AI systems poses significant risks. Discover the three major AI threat vectors including algorithmic attacks, physical vulnerabilities, and side-channel exploits that compromise edge AI deployments. Learn through real-world case studies, including adversarial attacks on facial recognition systems that demonstrate how malicious actors can manipulate AI behavior. Understand the complex balance between cost efficiency and security implementation in edge AI solutions, where resource constraints often conflict with robust protection measures. Examine emerging risks in agentic AI systems and their unique security challenges as autonomous agents become more prevalent. Prepare for the post-quantum computing era and its implications for current AI security frameworks. Gain insights into building proactive defense strategies that prioritize trust, transparency, and comprehensive security measures essential for the future of edge AI deployment and development.
