Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore advanced AI security vulnerabilities in this 40-minute Black Hat USA 2025 conference talk that demonstrates how artificial intelligence systems can be exploited to compromise enterprise environments without traditional attack vectors. Learn how attackers can leverage AI assistants to gain initial access, maintain persistence, harvest data, and exfiltrate information using zero-click exploit methods that require no credentials, phishing, social engineering, or human interaction. Discover real-world vulnerability chains affecting major enterprise AI platforms including ChatGPT, Gemini, Microsoft Copilot, Einstein, and custom AI agents that have evolved from simple assistants into powerful tools capable of reading search history, emails, and chat messages while manipulating enterprise environments. Examine the fundamental security challenge of prompt injection attacks and understand why they represent a manageable security problem rather than a fixable bug. Master the GenAI Attack Matrix security framework designed to help organizations protect against AI-based threats, compare mitigation strategies from various AI vendors, and analyze which approaches successfully prevent zero-click attacks. Gain insights into attack detection and mitigation by studying the breakdown of sophisticated AI exploits into basic tactics, techniques, and procedures (TTPs) presented by cybersecurity experts Michael Bargury and Tamir Ishay Sharbat from Zenity.
