Between September and October, large-scale cyberattacks are believed to have targeted Japan and South Korea, with suspicions that Chinese AI software may have been used in the attacks. Experts point to the increasing fragmentation of cybercriminal groups, which is likely to make it increasingly difficult to assess the full scope of the threat. Attacks Spread Across Multiple Sectors The incidents began coming to light in late September, when car-sharing service Times Car announced that personal information, including driver’s license details, had been leaked. Similar incidents subsequently affected major media organizations, critical infrastructure providers in the logistics and financial sectors, and companies providing everyday services, including department stores and major restaurant chains.
Government agencies involved in cyber defense were also targeted, including the Digital Agency and prefectural police departments in Okayama, Kanagawa, and Fukuoka. In South Korea, financial institutions, including major banks, were also hit. Chinese AI Tools Suspected A cybersecurity company affiliated with the GMO Internet Group found that a Chinese AI software tool called ARTEX had been exploited in attacks against Japanese companies. US cybersecurity giant CrowdStrike has also reported that cyberattacks in South Korea were directed using instructions written in Chinese.
Identifying the attackers remains extremely difficult, as they attempt to cover their tracks by routing attacks through electronic devices around the world. Generative AI Lowers Language Barriers Ani Terada, co-founder and CEO of AironWorks, a Tokyo-based company specializing in corporate security systems, says Japan is increasingly becoming a target of cyberattacks. As of October 6, a total of 119 information leakage incidents had been publicly disclosed in Japan that year. Of these, 81, or approximately 70%, came to light in July or later. Terada estimates that “the total for the full year will be roughly double last year’s figure of 84 cases.”
He says the increase since the summer has been particularly pronounced compared with trends in other countries. “One major factor is that generative AI has eliminated the language barrier when targeting Japanese speakers,” Terada explains. With generative AI, virtually anyone can now produce natural-sounding Japanese text, making it easier for attackers to send convincing phishing emails and lure victims to fraudulent websites. Japan may also be particularly vulnerable to attacks by China- and Russia-linked actors, given its strained political relations with those countries. In addition, the Japanese government has only recently introduced legislation on active cyber defense, which aims to prevent damage from cyberattacks through proactive measures.
The law took effect on October 1, and experts say attackers may have deliberately timed their campaigns to coincide with its implementation. Japanese IDs Command High Prices Weaknesses in Japanese companies’ organizational cybersecurity measures have also contributed to the scale of the damage. Many companies reportedly allocate a smaller share of their business spending to cybersecurity than their counterparts in the United States and other countries. Only some appoint dedicated executives, such as chief information security officers (CISOs). “Japanese companies are lagging behind in their organizational approach to cybersecurity,” Terada says, contrasting the situation with that in the United States, where approximately 90% of companies have executives in such positions.
On underground markets, personal information from countries with high GDPs tends to command higher prices. Japanese citizens’ identity documents are reportedly traded for as much as ten times the usual market price. Cybercrime Using AI Criminals are also diversifying the ways they profit from these intrusions. Their methods include extorting ransom payments by threatening victims, trading personal information, and using stolen data to commit fraud. Advances in AI are making these techniques increasingly sophisticated.
Terada warns that “voice impersonation attacks, in which an attacker makes a silent phone call and uses AI to learn from people’s voices are also beginning to spread around the world.” Even a word or two of audio recorded during a phone call may be enough to generate a synthetic voice that can be used to commit fraud, he says. “AI is enabling attackers to divide responsibilities among themselves, allowing even small groups to carry out these crimes,” Terada says, predicting that tracking down the perpetrators will become increasingly difficult. He also warns that cyberattacks tend to increase toward the end of the year, when consumer spending picks up, and urges both individuals and businesses to remain vigilant.
Culled from https://japan-forward.com/
