AI in cybersecurity is not a new concept. Almost every security vendor has been working on it for years. In fact, AI was a godsend for cybersecurity.
In the olden days, like 5 years ago, you pretty much needed to know what a threat was, what it did – what it’s “signature” was to be able to detect it and deal with it.
Vendors set up intricate traps – honey pots – they monitored traffic, did everything to stay on top of all the threats.  And frankly, they did a pretty good job of it.
But the sheer volume of new attacks and methods is simply overwhelming. According to a Forrester Research report I found online for 2019, 80% of cybersecurity decision-makers expected AI to increase the scale and speed of attacks and 66% expected AI “to conduct attacks that no human could conceive of.”

Well, if you listen to my sister podcast, CyberSecurity Today, that’s one prediction that came true. The host, my colleague Howard Solomon has no problem finding new threats to talk about – and he goes to air four times a week.
So at one point, this idea that we can know everything that’s out there and detect it breaks down.
And then there’s another problem – the sheer volume of attacks is astonishing. 20 years ago, a security expert at one of the major banks told me that if their firewall went down for 10 minutes, they’d be overwhelmed with attacks. Can you imagine what it’s like today?
And not just volume, it’s the speed of the attacks. If they get in to your system, they often take their time and set up an attack, spreading throughout your network, so that when they do mount the attack, it’s massive, fast and overwhelming.
People can’t move or think that fast.