top of page

Contrarian Views : AI and ML are no silver bullets for cybersecurity

In an earlier posting, we wrote about the promises of 5G for cybersecurity. It is supposedly made possible because 5G holds the promise that it will enable and speed up R&D into AI and ML.

Today we provide a contrarian view on AI and ML on why it might not be a 'cure all' for cybersecurity as indicated by expert opinions.

The basis for our contrarian views are that there are an increasing lack of skilled technical talents (easily confirmed and verified from lots of surveys taken to estimate the demand and supply of skilled cybersecurity professionals) and a predominant trend to cut application development costs producing skyrocketing number of insecure and vulnerable applications across.

Some or maybe a majority of the mid to smaller size organisations compete in an aggressive and competitive environment, resulting in them disregarding application security and privacy, which could also be possibly due to a lack of funding and other available resources.

AI is supposedly capable of learning and solving any set of diverse problems while ML is helpful for intelligent automation but real world applications for these appears to more theoretical than practical and there are limited literature on how cybersecurity benefits on a practical level.

Artificial Intelligence is supposedly capable of learning and solving any set of diverse problems while Machine Learning is helpful for intelligent automation.

Opinions are already being expressed about the ability of ML to handle and deal with the issue 'Rights to be Forgotten' - the rights of the individuals to ask any organisation who collects or kept data on the individual for their details to be removed or deleted from the organisation’s database. Some of the larger organisations - e.g., when you call some of these organisations, the pre-recorded voice will tell you that, your calls are being monitored for training purposes - kept these personal information of their customers separately and it will prove to be a challenge just to remove all these disparate set of similar data as and when an individual request for their details to be removed or deleted.

Concurrently, when AI and ML are being argued for improving cybersecurity, bad boys are not going to just sit down and rest. They would also be actively exploring and using ML to better profile victims and how to accelerate and camouflage attacks.

So the contrarian views - are the benefits of AI and ML for cybersecurity largely exaggerated? 

Hmmmmmm... let's wait for new developments or maybe a 3rd view which is neither for or contrarian?

Bình luận

bottom of page