Jul 16, 2026 · 58 min · 16 segments
AI is giving cybercriminals capabilities once reserved for elite hackers and powerful nation-states. Cybersecurity expert Nicole Perlroth joins Kara to…
Nicole PerlrothGuest
Kara SwisherHost
When you were on the show just over a year ago, you said you were most concerned about, quote, "Everything, everywhere, all at once cyber scenario or multiple simultaneous cyber attacks on critical infrastructure.
In a short period of time, we've experienced more geopolitical instability and rapid advances in AI, but so far we've avoided this worst case scenario." I- is that still your number one worry? Talk a little bit about what's changed in a year.
You know, last year I was saying we were worried about this quote, you know, "Everything, everywhere, all at once cyber scenario" based on what we were actually seeing Chinese hackers pre-positioning, which, to jog everyone's memory, and it's still happening, is water, power, pipeline infrastructure, high-speed rail, aviation, ports, logistics.
So we had to contemplate this, you know, nightmare scenario of what would happen if all of these hackers sort of detonated on the access they have all at once.
So it wouldn't just be one colonial pipeline, but, you know, hacks on critical infrastructure across the country.
I always thought that it was unlikely that they would detonate on all these things at once unless they were making a move on Taiwan, and they wanted us to think twice about whether we wanted to support this island 7,000 miles away, and that even in that case, they would maybe go for, you know, one water treatment facility and a power grid hack here or there, just flicker the lights on and off.
But as a byproduct, it can exploit code, which means it can find vulnerabilities in every major operating system, every browser, and chain them together into an exploit, so it can do things like...
We have found exploits using this model in most major operating systems, most browsers, um, et cetera, et cetera, et cetera." And these are capabilities that I, you know, the expert, as a human, would never have found otherwise.
And so the problem is Now these models can do what only previously the tier one elite nation state intelligence or their contractors could do
Therefore, and may not have done, just they're just using it as a strategic advantage versus they would do it.
And the people that can now do it are kind of crazy people possibly, or just chaos monkeys or whatever.
When you were on the show just over a year ago, you said you were most concerned about, quote, "Everything, everywhere, all at once cyber scenario or multiple simultaneous cyber attacks on critical infrastructure.
In a short period of time, we've experienced more geopolitical instability and rapid advances in AI, but so far we've avoided this worst case scenario." I- is that still your number one worry? Talk a little bit about what's changed in a year.
You know, last year I was saying we were worried about this quote, you know, "Everything, everywhere, all at once cyber scenario" based on what we were actually seeing Chinese hackers pre-positioning, which, to jog everyone's memory, and it's still happening, is water, power, pipeline infrastructure, high-speed rail, aviation, ports, logistics.
So we had to contemplate this, you know, nightmare scenario of what would happen if all of these hackers sort of detonated on the access they have all at once.
So it wouldn't just be one colonial pipeline, but, you know, hacks on critical infrastructure across the country.
I always thought that it was unlikely that they would detonate on all these things at once unless they were making a move on Taiwan, and they wanted us to think twice about whether we wanted to support this island 7,000 miles away, and that even in that case, they would maybe go for, you know, one water treatment facility and a power grid hack here or there, just flicker the lights on and off.
But as a byproduct, it can exploit code, which means it can find vulnerabilities in every major operating system, every browser, and chain them together into an exploit, so it can do things like...
We have found exploits using this model in most major operating systems, most browsers, um, et cetera, et cetera, et cetera." And these are capabilities that I, you know, the expert, as a human, would never have found otherwise.
And so the problem is Now these models can do what only previously the tier one elite nation state intelligence or their contractors could do
Therefore, and may not have done, just they're just using it as a strategic advantage versus they would do it.
And the people that can now do it are kind of crazy people possibly, or just chaos monkeys or whatever.
Every episode on Radar is fully transcribed, speaker-labeled, and rich with metadata. Here is a taste of this one. Try Radar for free to see the rest.
3 of 7
AI Finds Zero-Day Exploits
Discover AI locating hacks nobody else could
Crypto Heists Fuel Nukes
North Korea’s crypto jackpot powers its weapons
160K Job Applications Scheme
One cell blitzed 160,000 United States applications
+4 more clips · 8 min 15 sec of audio in all
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
All 7 clips — the highlight moments, each cut as its own audio, with a title and a speaker
All 16 segments — the transcript broken into labeled sections, every ad read marked
All 24 topics — jump to every other episode discussing the same subject
Every related episode — other shows Radar links to this one
No account is needed to search Radar.
