By using the new AI technology, countermeasures can speedily be put in place for cyberattacks, where action is necessary
Fujitsu Laboratories Ltd., a fully-owned subsidiary of Fujitsu Ltd and a leading research center in the world, announced that it has developed an artificial intelligence (AI) technology that automatically determines the action that needs to be taken in the case of a cyberattack.
In case of a cyberattack on the business network, several security appliances detect the hit on the servers and devices of the network. Traditionally, a specialist in cyberattack analysis then manually checks and investigates the degree of threat, in order to determine the action to be taken to minimise the damage from such attack.
In order to safeguard the necessary training data required to create a highly accurate AI technology, Fujitsu Laboratories developed a technology that detects and extracts attack logs that display the behaviour of a cyberattack, from vast amounts of operations logs. In addition, it created a technology that expands on small number of training data extracted in a way which does not spoil characteristics of an attack. This creates a considerable amount of training data.
Quick response in case of cyberattack
In simulations using these technologies, a match rate of around 95 per cent has been achieved in comparison with specialist’s conclusions with regard to the need for action. Besides, they did not miss any attack cases that needed a response. Apart from this, the time required to reach a conclusion was cut down from several hours to some minutes.
By using the new AI technology, countermeasures can speedily be put in place for cyberattacks where action is necessary, thus leading to business continuity and prevention of loss.
Moreover, by combining the latest technology with Fujitsu’s high-speed forensic technology, the response sequence, from analysis of cyberattack to the instructions for action, can be automated, thus enabling quick response to cyberattack and minimising the damage.