It’s always good to take the long view – and when it comes to cybersecurity, Cato Networks’ co-founder and CEO Shlomo Kramer has a longer view than most. His 30-plus years in the sector have spanned the early days of the web through to today’s AI-fueled free-for-all. Indeed, Kramer recalls the dawn of the internet being greeted as an “overnight revolution” by businesses. “Organisations simply were afraid to connect to the internet,” he tells Tech Monitor, recalling his days helming Check Point Security. The emergence of core cybersecurity principles, meanwhile, allowed them to thread the needle between their own protection and building an online presence.
Founding his first startup shortly after leaving the Israeli Defence Forces (IDF) in 1993, Kramer’s innovative firewall technology helped pave the way for thousands of businesses to confidently carve out online niches. The Israeli entrepreneur has since established two more household names in cybersecurity – Imperva and Cato Networks – and is now celebrating the adoption of the latter’s SASE Cloud Platform by over 3,000 businesses to date. In the following interview, edited for length and clarity, Kramer talks to Tech Monitor about the longue durée of cybersecurity, and how recent developments in the threat environment not only merit continued innovation from IT defenders but also a great deal more attention on cybercrime from governments.
Tech Monitor: You’ve talked previously about three generations of IT security: software; appliances; zero trust network access. IT security spending is going up. But are we any more secure?
Shlomo Kramer: The cost is going up, and organisations are less secure. The cost of data breaches alone has grown by more than 40%. And the reason for that is the level of complexity in the organisations impacted. They’re too clunky. In a mid-sized company, the CIO has tens of products that they need to integrate. Essentially, IT security has not been digitally transformed, so it can’t address the needs of a digitally transformed organisation in a way that is operationally efficient enough and business agile enough.
So, there was a need to digitally transform network security, which was our slice, and that’s what we’ve done. We’ve created a cloud network with network security embedded inside. We have more than 85 POPs [points of presence] today around the world, all running the same network and network security stack. You can on-ramp to that network using SD WAN and ZTNA and IPsec, and that network provides all the security requirements of an organisation. Other vendors are more, I would say, appliance vendors, attempting to protect their turf, and we’ve built a platform.
How is AI affecting your world? Does everybody get it?
[AI] is obviously a much bigger innovation than anything. It’s both a tool for attackers to create phishing emails and malware and a shield for the defending party. We always used AI.AI depends, first of all, on data. We have a huge data lake in the cloud from all of our 3,000 plus customers, the metadata, where we hunt for new threats, where we find unusual behaviours, where we test silently signatures, and where we build AI algorithms that are accurate enough to put in line.
From an operational perspective, we have Copilots. Our XDR folks, just as an example, can tell you exactly what happened in English, and miraculously, are able to find the organising principle of the incident. That’s mind-blowing. If you look at the security events, you can search in plain English. You don’t need to query. Obviously, the organisation itself, the support, and every aspect of the organisation is driven by AI today.
But who’s moving faster when it comes to leveraging AI? The bad guys or the good guys?
I will point you to our own CATO CTRL Research, where you will see amazing AI attacks, where you manipulate the model to do stuff that is cutting-edge in terms of generating malware.
I’m afraid that the most dangerous side of AI is on the fraud side, and kind of telling truth from lies and it has serious consequences for society.
So, beyond what’s happening with AI, what is still broken in cybersecurity?
If you’re a citizen or a corporation and somebody sends a missile on your house or your factory, your government is going to do something about it. If threat actors aim a cyberattack at you, perhaps you can call…I don’t know who, the FBI or another agency, depending on the country? And perhaps they’re going to write it down and do something about it. Perhaps not.
I think that needs to change. That’s in times of peace. What happens in times of war? What happens if there’s a mass attack on that segment of the market that is really not ready at all? I think there needs to be some more government involvement in that in this part of the warfare on the defensive side.
Cybersecurity is like a pharmaceutical company that has, for 30 years, focused on developing the best molecules in the world but forgotten to develop relevant delivery mechanisms for them. I think that the next innovation needs to really focus on how you deliver this in a way that is consumable and effective and agile and affordable to everyone, from the small organisation to the largest corporates.
Read more: Look for long-term value with generative AI and not short-term fixes, says Deutsche Bank’s Christoph Rabenseifner
