Expert Discusses the Evolution of Security in the Context of AI

In this new video from @CloudSecurityPodcast, Anker, a cybersecurity expert, shares his thoughts on the evolution of security in the context of artificial intelligence (AI). The discussion begins with an introduction to Anker's journey in the field of technology and security, highlighting his experience in creating startups and his work at Palo Alto, Redlock, and Prisma Cloud. Anker explains that AI is fundamentally changing the way applications are built and secured. Unlike traditional systems, AI enables the creation and consumption of software using natural language, paving the way for an explosion of applications. However, this new approach also presents unique security challenges. For example, attackers can use simple natural language-based techniques to exploit systems, making traditional security methods ineffective. A key point of the discussion is the difference between native AI applications and traditional SaaS applications that integrate AI features. Anker categorizes companies into three groups: AI natives who build applications from scratch with AI, AI immigrants who add AI features to their existing applications, and AI explorers who are trying to become AI natives. He emphasizes that native AI applications, like Cursor and Glean, completely rethink the application stack, including infrastructure, data, model, and query layers. Anker stresses that security needs to be rethought to adapt to these new architectures. For instance, data security and data pipeline security in rack systems differ from security at the inference layer against data leaks and evasions. He introduces the concept of "native AI guardrails" to protect against these new threats, highlighting that only AI can secure AI. The discussion also covers practical implications for security teams. Anker recommends starting by understanding the use of AI within the organization and implementing guardrails to secure interactions with AI. He predicts that the AI threat landscape will be more significant than all other security categories combined, affecting application, identity, and endpoint security. In conclusion, Anker shares his thoughts on the future of security in an AI-dominated world. He encourages security teams to train and adapt to these new technologies while recognizing the challenges and opportunities they present. The video ends on an optimistic note, emphasizing that despite the challenges, we are living in an exciting time where AI can be a powerful enabler for building the future. To learn more, watch the full video at the following address: https://www.youtube.com/watch?v=9AWJ09_XNGI