In today’s world, speed, agility and scalability are essential for organizations and businesses if they want to become successful and stay relevant. On-premises IT can’t provide them with the speed, agility and scalability cloud environments can, so the continued embrace of cloud is inevitable.
Unfortunately, the same characteristics – speed, agility and scalability – also apply to the bad guys. We now see, for example:
These developments signify a serious need to change the approach to securing organizations.
Effective security can no longer rely on a point product approach, for which the acquisition, implementation and training might take weeks or even months. In the cloud era, that’s no longer a viable tactic because the manual use of these point products makes organizations slow and reactive. In other words, we simply cannot defend our organizations against highly sophisticated, automated and agile threats by using old-fashioned, non-automated and non-integrated security.
Cybersecurity technology companies understand this and have for some years been investing in cloud computing, including ways to secure cloud environments and deliver security via cloud-based services. An example of a cloud-delivered security service is a threat intelligence capability in the cloud, which uses the speed and scalability of the cloud model for its software analysis process and can deliver the protection needed within a very short time frame.
The core of what will make cloud computing capabilities continually useful is big data analytics. Without big data analytics, it’s impossible to apply machine learning, which is essential for automation and the required speed of operations. Unfortunately, the terms ‘big data analytics’, ‘machine learning’ and ‘artificial intelligence’ are often confused and used interchangeably. Several cybersecurity companies claim to use artificial intelligence for their services, but they probably mean big data analytics and machine learning. To explain this in simple words, here are the definitions I use to clarify these terms:
How are big data analytics, machine learning, artificial intelligence or the combination of these capabilities best used to protect organizations from cyberattacks?
Unfortunately, there’s no silver bullet yet in this context, although large amounts of data can be better and more quickly handled by machines than by humans (see the threat intelligence example above). The challenge is that AI, especially, is being over-marketed for cybersecurity, but the technology has its limitations: AI is never designed to work in adversarial environments. It works quite well in games like chess or go, where the rules are well-defined and deterministic.[4] But in cybersecurity, these rules don’t apply, and the ‘bad guys’ are constantly evolving and adapting their techniques. At this moment, AI is less suitable because it cannot adapt to the fast and unpredictable environment. This will no doubt improve in the future.
Analyzing data kept in one place also means that place is a single point of failure. An attacker only needs to make subtle, almost unnoticeable changes to the data in this one data location, which could undermine the way an AI algorithm works.[5] Therefore, it’s essential to understand how big data analytics, machine learning and AI work; recognize the limitations; and act accordingly, not on hype.
In today’s world, the use of big data analytics, machine learning and AI provides several advantages in the cybersecurity domain – especially in the threat intelligence, behavioral analytics and cyber forensics areas – but there’s still a long way to go before we can completely rely on these capabilities in cybersecurity. When we get them right, we will truly maximize our investments in cloud.
By submitting this form, you agree to our Terms of Use and acknowledge our Privacy Statement. Please look for a confirmation email from us. If you don't receive it in the next 10 minutes, please check your spam folder.