{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"Watters Edge","title":"Black Hat 2026 - Watters Edge - Episode #5","html":"","width":"100%","height":180,"duration":917,"description":"AI was the word of the week at Black Hat 2026, and this episode captures that in real time through a supercut of quick conversations recorded live on the floor. Voices from across the industry, including Cisco, SAP, Armadin, ICM Cyber, and former NSA leadership, weigh in on what stood out most this year, how they see the third wave of cybersecurity taking shape, and what they think the industry gets wrong.\n\nA clear theme runs through nearly every answer: AI is no longer a talking point, it is the water everyone is swimming in, for better and for worse. Between the excitement about agentic AI and pointed calls to get back to security basics, this episode captures where the industry's head is really at heading into the next wave.\n \nTakeaways:\nChad Skipper (Cisco) points out that frontier and agentic AI models need real security guardrails, since the same systems that make them powerful also make them capable of acting unpredictably without oversight.\nDonna Estren notes that AI is now touching every corner of security, from OT and supply chain to endpoint and DevSecOps, meaning point solutions increasingly need to think about AI's ripple effects across the whole stack.\nGreg Davidson (Armadin) makes the case that the smart move for security leaders right now is narrowing focus onto what is actually exploitable rather than trying to chase every vulnerability with a limited budget.\nKen Foster (ICM Cyber) argues that agentic AI is most valuable when it is used to finally fix the fundamentals the industry has struggled with for decades, not just as a buzzword layered on top of old tools.\nRob Joyce (former NSA) warns that most organizations still underestimate how fast offensive AI tooling is advancing, and that doing the basics well remains the single biggest lever for actual security.\nMultiple speakers flagged the same myth from different angles: neither compliance nor threat modeling actually solves security problems on their own. They are tools to help manage risk and...","thumbnail_url":"https://img.transistorcdn.com/jNADFOLa2hUu-KALPMSZ3sOeDtfDXDIggtDJVeKDZfg/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS8xNjZj/YTM0ZGYwZjc0Nzhi/NWMyYjBiZDUxMzU3/YWI2Ny5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}