AI, Identity & the Nature of Knowledge
Combined Session
Thursday, May 08, 2025 15:35—16:35
Location: B 05
Log in to download presentations
Thursday, May 08, 2025 15:35—16:35
Location: B 05
Watch the video
Deepfake technology is advancing rapidly, making traditional identity verification methods increasingly unreliable. The arms race between deepfake generators and detection technologies is escalating, making it costly and unsustainable for organizations to rely solely on AI-driven detection tools.
Instead, intelligence agencies have long relied on low-tech but highly effective verification methods—techniques that could prove invaluable in today’s cybersecurity landscape. This session explores how espionage tradecraft, including signs, countersigns, and challenge-response techniques, can be adapted to modern cybersecurity to verify identities in an era of AI-driven deception.
Using real-world cases like the $25 million CFO deepfake scam, we will discuss why organizations must go beyond technical solutions to combat impersonation threats. Attendees will gain practical strategies for integrating human-centric authentication techniques into their identity verification processes.
Watch the video
As AI agents become integral to business processes and systems, ensuring they operate securely and within defined boundaries is critical. Much like human users, AI agents require robust IAM solutions to control their interactions with sensitive data and systems.
We’ll explore the reasons why AI agents need IAM access controls, examine potential risks of unregulated access, and discuss best practices for integrating IAM strategies into AI workflows. By treating AI agents as entities with specific access rights, organizations can mitigate security threats, ensure compliance, and improve trust in AI-powered systems.
Join us as we explore how IAM can safeguard AI systems in today’s interconnected world.
Watch the video
The question of how machines perceive, interpret, and understand the world challenges our traditional concepts of knowledge. This session explores the philosophical foundations of knowledge, drawing on Immanuel Kant’s Critique of Pure Reason to examine the boundaries of human and machine cognition. Kant famously argued that our understanding of the world is shaped by innate structures of thought, which mediate our experience of phenomena and prevent us from directly accessing the "thing-in-itself" (noumena).
In parallel, AI systems process data through algorithms and predefined structures, interpreting information in ways that are both similar to and different from human cognition. How do Kant’s epistemological conclusions about the limits of knowledge apply to AI’s capabilities? Can AI transcend the boundaries of human understanding, or does it merely replicate the cognitive constraints of its creators?
