Summary
On the All-In Podcast, IBM’s Arvind Krishna and Palo Alto Networks’ Nikesh Arora discuss how AI is changing cybersecurity, the looming quantum threat to encryption, and what it all means for enterprise software.
Krishna starts with quantum computing, describing DARPA’s Quantum Benchmarking Initiative, where a couple of hundred companies have narrowed to roughly 13. He sketches the milestones to watch, such as demonstrations of error correction and scaling from around 10,000 gate operations toward millions, and uses 2030 as a rough midpoint for when today’s cryptography could be broken. He warns that stolen encrypted data could be cracked years later and says post-quantum cryptography is well understood but under-adopted.
The conversation then turns to AI-powered cyber warfare. The guests explain why attackers have the advantage, since they need only one vulnerability while defenders must find and patch all of them. They note that models like Mythos surface real flaws but still hallucinate a meaningful share, so findings need verification, and that AI-written code is not error-free. They also discuss pooling vulnerability data among trusted parties and the debate between open and frontier models. Business leaders and security professionals will find a candid look at where AI and security risk are heading.
📺 Source: All-In Podcast · Published October 21, 2026
🏷️ Format: Podcast
