TOKYO -- Cyberattackers are launching increasingly sophisticated strikes around the world, but many companies have all too often been too slow in detecting and taking action against new tactics used by black hat hackers.
The gap between attacks and detection and response has prompted Japan's NTT Communications and SoftBank to both start planning new artificial intelligence-based services that can instantly protect companies from potential cyberattacks.
Many big businesses have not only installed anti-virus software onto PCs, but also introduced surveillance solutions provided by computer security specialists in a bid to fight increasingly subtle malicious exploitation of computer systems by unauthorized users.
Through big data analysis, these services can screen illicit operations such as transmission of confidential information to external receivers and shut down communication upon detection. However it usually takes 8-15 minutes, even for a security expert, to identify a new mode of attack as some of the process has to be done manually.
NTT Communication's new service, which will be launched worldwide next spring, can automatically identify patterns in a virus and the modus operandi of previous electronic crimes, and detect new ones by analogy through analysis of large amounts of data.
If an attacker automatically changes the URL to which they intend to send confidential information, the solution would still be able to recognize it instantly. Reportedly, the service will be able to detect over 99% of unauthorized accesses.
The automatic detection function is also expected to assist cybersecurity experts who are significantly understaffed in the face of increased cyberattacks.
NTT Communications' service is unique in its ability to protect a variety of forms of intellectual property, such as customer information and design drawings. So far, AI-based security measures have only been used for certain purposes, such as the detection of fraudulent online financial transactions. It will be the first commercial service to have such a wide range of applications.
The service will be available from around 100,000 yen ($824) per month. The company hopes to earn 20 billion yen per year from the service, including add-on services.
SoftBank, meanwhile, has recently injected around five to six billion yen in Cybereason, an AI-based anti-virus software developer based in Boston, Massachusetts.
The Japanese telecommunication giant's plan is to launch a monitoring service that utilizes Cybereason's solution as early as next year. The offering is expected to even be able to detect unknown viruses that invade a company's internal system, by tapping into knowledge acquired and accumulated through experimental learning. The solution is expected to help companies reduce information leaks. Fees for the service have yet to be finalized.
Cybereason is a new venture company founded by former Israel Defense Forces security experts, rumored to possess state of the art information security technology. The company has also received capital injection from such companies as leading American defense contractor Lockheed Martin.
According to National Institute of Information and Communications Technology, Japan's sole national research and development agency specializing in the field, the number of cyberattacks on government agencies and private companies in Japan grew by 100% from 2013 to roughly 25.6 billion attacks in 2014.
Dramatic technological developments in the 1980s enabled AI to make case-by-case decisions in relatively general fields such as medicine and finance. However, there were still many hurdles to clear before applying it to specialized fields, such as the necessity to specify precise requirements after having it learn data structured by human beings.
The launch of these services is owed largely to the development in the early 2000s of information processing technology that can simulate the reasoning process of the human brain, which has enabled AI technologies to independently identify complex patterns hidden in vast amounts of data.