Show Notes
Keren Katz exposes novel risks posed by GenAI and agentic AI while reflecting on unintended malfeasance, surprisingly common insider threats and weak security postures.
Keren and Kimberly discuss threats amplified by agentic AI; self-inflicted exposures observed in Fortune 500 companies; normalizing risky behavior; unintentional threats; non-determinism as a risk; users as an attack vector; the OWASP State of Agentic AI and Governance report; ransomware 2025; mapping use cases and user intent; preemptive security postures; agentic behavior analysis; proactive AI/agentic security policies and incident response plans.
Keren Katz is Senior Group Manager of Threat Research, Product Management and AI at Tenable, a contributor at both the Open Worldwide Application Security Project (OWASP) and Forbes. Keren is a global leader in AI and cybersecurity, specializing in Generative AI threat detection.
Related Resources Article: The Silent Breach: Why Agentic AI Demands New Oversight State of Agentic AI Security and Governance (whitepaper): https://genai.owasp.org/resource/state-of-agentic-ai-security-and-governance-1-0/ The LLM Top 10: https://genai.owasp.org/llm-top-10/ A transcript of this episode is here.
More from Pondering AI
Colin Levy provides counsel on cultivating judgment, acknowledging the human factor and minding more than the tech when adjudicating the governance and use of legal AI.
Cecilia Dones guides leaders toward meaningful AI adoption by remaining endlessly curious, prioritizing social connection, valuing deviation, and embracing uncertainty.
Kelly Monahan reimagines the story of AI at work, helping leaders to reclaim the plot and draw up new business blueprints that optimize human capability and innovation.
Leigh Felton believes we can benefit from AI systems without surrendering to them if we carefully consider the environments and behaviors we promote while deploying AI. Kimberly and Leigh discuss the power of perspective; organizations as people; using AI consciously; the AI environment and library etiquette; how AI use shapes human behavior; whether AI systems are adaptive; AI authority and narrow optimization; the gap between lived experience and historical data; the AI Emperor’s new clothes; quiet organizational failures; and who benefits from humanizing AI. Leigh Felton is the President of the AI for Job Security Foundation. Leigh applies her extensive experience leading communications, business strategy and operations at the world’s largest tech companies to help organizations make conscious decisions about and with AI. Related Resources: I Could’ve Been CEO Yesterday V1: But AI Algorithms Say I’m a DEI Risk Today (book) I Could’ve Been CEO Yesterday V2: Does AI Decide Who’ll Be CEO Tomorrow? (book) Inside Power w/ Leigh Felton (newsletter) A transcript of this episode is here.
Sarah Barrington shifts from simply asking “real or fake?” to deeply considering the ways AI and synthetic media influence our notions of culture, identity and trust. Kimberly and Sarah discuss her road from the fast track of F1 to the frontlines of AI; the genesis of photography; supercharging existing harms with GenAI; the unsavory roots of synthetic media; microtargeting and information warfare; the liar’s dividend; content detection and digital identity; the agentic AI challenge; culture, identity and relevance; Mythos, AI access and proliferation; anti-regulatory narratives; the nuclear analogy; equipping policy makers and the public; questioning everything; and true AI enablement. Sarah Barrington is an AI researcher and PhD Candidate at UC Berkeley where she is part of Professor Hany Farid’s digital forensics lab. She also holds multiple cybersecurity positions. A specialist in generative AI & deepfake analysis, Sarah’s work has been cited at the Nobel Prize Summit, on NPR, by the UN and in numerous publications. Related Resources: David Attenborough and the Voice That Revealed a Planet (article) Sarah’s Research and Publications (website) The DeepSpeak Dataset Get Real Security Labs A transcript of this episode is here.