AWS Startups Blog

Darktrace’s AI-based Cyber Security System Enables Companies to Spot and Stop Cyber Threats

In the past 5-to-10 years, data breaches at major entities like Target, Adobe, LinkedIn, and Equifax have frequently made headlines. These breaches led to millions in legal fees and settlements, but the damage doesn’t stop there. The amount of time and resources companies need to rebuild their brand image after a major security incident can be staggering.

One of the main problems is that attackers are getting more sophisticated by the day. Enterprises large and small are working just as fast to keep up, but attacks are coming through channels where you’d least expect them. More important, it’s becoming increasingly difficult to identify a threat in its early stages, especially when you have a massive infrastructure to monitor.

That’s where Darktrace comes in. Founded in 2013 by mathematicians from the University of Cambridge, Darktrace has developed an AI-based system for cyber security designed for the modern age.

Mike Beck, Global Head of Threat AnalysisMike Beck, Global Head of Threat Analysis at Darktrace explains how their approach is similar to the human immune system. “We have skin that defends us from a whole range of stuff entering our bodies, but there’s a reality to the fact that some things get through. And as they enter the body and start to move around and latch onto things, an immune system response kicks into gear.”

Darktrace’s platform works in a similar way. Its flagship offering, the Enterprise Immune System, is a self-learning technology that detects threats and insider attacks at the earliest stages. The Enterprise Immune System continuously learns your entire business footprint and understands what normal operations look like, so it can pinpoint suspicious behaviors when they arise.

“For us it’s about latching onto things as they start to cause you pain and being able to surface that in a meaningful way to the security team to either take action using incident response processes or by autonomously using Darktrace itself.” says Beck.

That is handled by Darktrace’s Antigena, an autonomous response solution. Antigena responds to cyber threats in seconds, which buys security and technology teams precious time and frees up valuable resources. If you think of the Enterprise Immune System as your business’s immune system, Antigena would be a sort of digital antibody—attacking threats across your entire ecosystem and using data to inform and modify its response.

One trend that the company has seen and capitalized on is the move to the cloud across various industries, per Beck.

“While some of our clients run on-prem, the majority of AI workloads nowadays are being done in the cloud. Because of this, we work closely with AWS both from an infrastructure perspective and in supporting our clients.”

As a leader in the cyber security industry, Darktrace specializes in protecting cloud environments, SaaS, email, and even IoT devices, so customers can rest easy knowing they’re secure, but the decision to take action and build out a robust cyber security program is ultimately up to leadership at vulnerable organizations. According to Beck, the most successful organizations are having cyber security conversations at the board level and thinking holistically about their approach. That increased focus has been a boon for Darktrace, which now counts over 3,500 companies as customers.

“Another trend we’re seeing are boards becoming savvier about digital transformation and the risk of cyber attack. More conversations are being had at that top level on how to manage such an event, which has led to improved cyber posture in many industries.” Beck says.

As for what’s on the horizon with Darktrace, there are no signs of slowing down. In fact, Beck and team launched a new offering to help customers that don’t have the resources to scale their security operations center, or SOC, teams.

“We released something quite recently called Cyber AI Analyst. This product sits above the unsupervised machine learning and it’s taking the output of those results, effectively giving customers the benefit of an AI analyst alongside them 24/7 – for our customers, this has resulted in a 92% reduction in time to triage. Scaling out SOC teams is not something that’s possible for many businesses, so having AI continually helping with that scale factor is critical.” says Beck.