The Tech Blog Writer Podcast - Inspired Tech Startup Stories & Interviews With Tech Leaders, Entrepreneurs And Innovators
Claroty on Combating Model Poisoning and Adversarial Prompts
- Autor: Vários
- Narrador: Vários
- Editor: Podcast
- Duración: 0:35:29
- Mas informaciones
Informações:
Sinopsis
AI is rapidly becoming part of the healthcare system, powering everything from diagnostic tools and medical devices to patient monitoring and hospital operations. But while the potential is extraordinary, the risks are equally stark. Many hospitals are adopting AI without the safeguards needed to protect patient safety, leaving critical systems exposed to threats that most in the sector have never faced before. In this episode of Tech Talks Daily, I speak with Ty Greenhalgh, Healthcare Industry Principal at Claroty, about why healthcare’s AI rush could come at a dangerous cost if security does not keep pace. Ty explains how novel threats like adversarial prompts, model poisoning, and decision manipulation could compromise clinical systems in ways that are very different from traditional cyberattacks. These are not just theoretical scenarios. AI-driven misinformation or manipulated diagnostics could directly impact patient care. We explore why the first step for hospitals is building a clear AI asset inventory