How advanced technology is associated with Cyber Security? (1/2)

This article is written by Mr. Kenji Kobayashi , Senior Fellow, Institute for International Socio-Economic Studies（NEC Group）

Visits to nations of huge cyber security power
In order to understand the latest situation how advanced technology such as AI (Artificial Intelligence) is utilized in the field of cyber security I visited in this June UK and Israel, well known as most powerful and advanced nations in cyber security.

Experts whom I interviewed at UK NCSC (The National Cyber Security Centre) commented “So far, AI (Machine learning (ML)) is with us” in response to how AI is utilized in the field of cyber security. NCSC is actually utilizing AI technology to process a huge amount of data and information collected in UK and AI is so far very helpful and useful.

On the other hand they have opinion that the malicious use of AI by cyber criminals are only limited to automated conventional cyber-attacks for the time being. One of the reasons behind is quoted that most of cyber criminals (organized group) are doing the actual economical business and already have more profitable, more effective and easier attacking ways than use of AI. They are not seeing AI worth utilizing from the economical point of view.
This deep insight for cyber criminals are based on the collaborated study of experts in a variety of sections (social economy, psychology, computing and so on) within NCSC.

This approach could be a role model for Japanese experts who tends to focus on the technological aspect.

Malicious use of AI
But most experts including NCSC predict that AI will be utilized with malicious purpose in the near future. Especially, they are concerned about cyber attacking for AI itself. For example the malfunction is caused by false data, and the training data is copied or identified (Adversarial attack). These attacks cannot be easily detected because of the “black-box” of machine learning.
Professor Alessio R.Lomuscio of Imperial College London has started studying the verification of ML system one and half year ago. He said that “there are only 3 academic papers and this theme is considerably challenging.”
It is very difficult for us to explain properly reason and to guarantee for the results of ML system. So, we have a lack of such methods and tools.
The professor gives warning message that “Current AI should not be utilized in the field where the safety is prioritized.”
It should be remarked that the practical use of AI should be made cautiously and carefully with proper understanding on the technical limit of AI.

Risk Taking
In order to prepare for future high level of sophisticated cyber-attacks, the coordinated research and development is progressed among government, military, private and academic sectors in nations of the huge cyber security power.
In southern part of Israel there is a city of Be’er Sheva where National Cyber Directory, Intelligent and IT units of National Military Defense, global IT companies, and academia are located in the same premise. Prime Minister Netanyahu stressed the importance of taking risk to be one step ahead to cyber criminals even if leakages of state confidential security information are inevitable.
In the facility (of Prof. Yuval Elovici at Ben-Gurion University) I visited, they are attacking and hacking the various IT devices by all possible measures, as a research of cyber physical security. It seems that they are all benefited with the actual and practical experience of intelligence units.

It is of regret that in Japan we are lacking in this sort of actual circumstances.
In the next report I would like to talk about importance of information-sharing and how Blockchain (Distributed ledger technology)　is utilized.