Nithen Naidoo dropped by the ITWeb studio for a conversation beyond beyond the potential of AI, how Snode uses AI in a way that helps organisations reduce complexity, focus on the risks that matter most and strengthen resilience across their digital environments.
#snode
Snode Technologies was using AI long before it became the industry’s favourite buzzword. Today, the Centurion-based cyber defence company protects around 8 million devices across six continents.
“In truth, we weren’t looking to leverage AI in a cyber security product. We were looking for a solution to a problem,” says Nithen Naidoo, founder and CEO of Snode. Signature-based detection could not keep pace with how quickly threats were evolving, and by 2017, that search had taken Snode from Bayesian mathematics into machine learning and deep neural networks.
AI now shapes both what Snode builds and how it builds. “Code has become something that is machined rather than a human process,” explains Naidoo. Writing code once dominated the development cycle, but now the team spends its time on design, quality assurance, security testing and AI red teaming.
“At Snode, we don’t believe AI is going to replace humans,” he adds. For Naidoo, keeping ‘humans in the loop’ when processes are AI-driven could be compared to intelligence amplification, the philosophy of using technology to enhance human cognitive capabilities rather than replacing them. “You’re looking at the synergy of humans doing what humans are good at and AI doing what AI is good at,” says Naidoo.
Because AI sits at the core of the platform rather than being retrofitted, Snode works with behaviours instead of isolated indicators like IP addresses and file hashes. “Your firewall is producing an alert, your endpoint is producing an alert, your ERP is producing an alert, but it’s actually one attack,” Naidoo explains. The analytical engine correlates that activity into a single, contextualised view of the attacker’s kill chain.
Snode’s digital twin then takes things further, simulating how active threat actors would move through an organisation’s real environment. “Imagine a machine that could run thousands of simulations in real-time and find out exactly how threat actors would attack you,” he says. Instead of drowning in a report listing a thousand critical vulnerabilities, security teams leave the room knowing precisely which exposures matter and what to fix first.
Nithen Naidoo, CEO, Snode.
Rather than relying on public large language models, Snode trains its own small language models, built on a decade of analyst and client data expanded through data synthesis. These are hyper-specific, lightweight and portable enough to run at the edge on IOT sensors, phones and even satellites.
“If it can’t pick up anti-virus signatures, it’s not going to use ChatGPT,” says Naidoo. Smaller models are also easier to trust. “Too much of AI is like a black box and it needs to be more of a glass box. You need to understand how the AI works in order to understand how it can be manipulated or misused,” he explains. And in a hyper-connected world, AI-native cyber defence is only as good as its ability to surface that understanding while a threat is still unfolding.
“Knowing about an attack five minutes after it happened is five minutes too late,” he advises. “You have to move at machine speed, because you’ve been attacked at machine speed.”
You can view the full video here.
* ITWebTV Biz episodes are sponsored.
