Nov 6, 2026 · 34 min
AI could turn cyberattacks into massively parallel campaigns
Keeping cyber threats on ice.
The episode connects real-world breaches and fraud with a looming shift in cyber defense, as unrestricted AI could scale attacks beyond human capacity.
- 1Industrial control systems, healthcare networks, and identity infrastructure remain exposed to attacks with physical and financial consequences.
- 2OpenAI’s GPT-6 Cyber demonstrates how capable security models may find vulnerabilities while requiring safeguards for authorized use.
- 3Expel CTO Greg Notch argues defenders must prepare for massively parallel attacks, not merely faster versions of today’s incidents.
Don't miss
Greg Notch warns that unrestricted AI could enable attackers to discover vulnerabilities and launch operations in parallel at a scale defenders cannot match.
The brief
The briefing moves from attacks on industrial refrigeration and cold-chain systems to breaches at Chainalysis and DigiCert, showing how cyber risk reaches food, medicine, credentials, and identity data.
OpenAI’s GPT-6 Cyber is presented as a highly capable security model that can find vulnerabilities and exploits, while D-Link switch flaws underscore the persistence of basic defensive gaps.
An alleged AI-enabled impersonation scheme used fabricated executives and attorneys to direct wire transfers, targeting thousands of companies and causing approximately $129 million in losses.
Expel CTO Greg Notch argues that frontier AI development will not slow and that unrestricted models could assist malware development, reverse engineering, and offensive cyber operations.
Notch’s central warning is about scale: defenders must build layered defenses for massively parallel attacks that could overwhelm personnel, budgets, and existing response models.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.