AI and Cybersecurity (1/2)Important issue associated with possible collapse of the nation

This article is written by Mr. Izumi Harada , Executive Fellow, Institute for International Socio-Economic Studies（NEC Group）

Attacking party is more advantageous
In the cyber space it is highly difficult to defend against unforeseeable cyber attacks created one after another. So attacking party is substantially advantageous to the defending party. Lately the defending party is, however, strengthening its defense power with advanced AI(Artificial Intelligence) technology.

With utilization of advanced AI technology it becomes possible to:
-collect the information and data on attribute and possible damage of malwares traded in the darkwave
-monitor a large amount of data communication in real time
-collect and store the related information on data, sender, connections commonly used in the attack

Having analyzed the information by machine learning and Neuro Linguistic Programing it becomes possible to predict possible new attack and to prepare how to deal with it.

Moreover AI works 24 hours a day and 365 days a year without interruption and lessen the work load of administrators with automatic programming.

Difficulty in collecting essential data
At present it is still not easy to collect essential data associated with attacks. There remains also the issue that it is difficult to remove internal and external factors to cause malfunction of AI which is gullible by nature.

It is, however, expected in the near future that defense power becomes stronger through the sophisticated autonomous learning functions, and by more intellectual, speedy and highly precise autonomous monitoring on the attacks by collecting the information associated with the relevant cyberattacks out of an enormous amount of open data. 

On the other hand AI is very useful to narrow down and specify the target to attack and replace gradually human work load by autonomous processing.

AI is capable of organizing sophisticated cyber attacks more speedy and effectively than what human hacker can do, and, is capable also of helping hackers to customize the attack. 

So far hackers have taken the measures of using multiple number of similar malwares with slightly different features to avoid the detection. 

Sooner or later it is expected that more and more new malwares are created by advanced autonomous malware development with the help of machine learning.

AI vs AI
At present it is considered still costly for cyber criminals to utilize AI technology and no new cyberattacks organized with real machine learning have been reported. 

It is, however, to be noted that lately systems to use machine learning algorithm are offered inexpensive and it is getting easier to access learning data. 

It is, therefore, becoming less and less difficult for cyber criminal to utilize AI technology for cyberattacks.  

In near future the cyber space will be the battle field of AI vs AI. 

The main battle field will be between cyber criminals and not only private corporations and entities but also national security and intelligence organizations. 

It is to be stressed that the development of AI in the field of cyber security is substantially important issue associated with possible collapse of the nation itself. 

It is understood that the development of AI technology has been well progress sed in the military field and AI is so well developed that it may change the conventional military system.  

In the next issue I will report the latest situation of AI military revolution.