TOKYO: (Bernama-Kyodo) Over 80 per cent of email scams with identifiable senders targeted Japanese recipients in May, as advances in generative artificial intelligence (AI) allow the use of more natural language, Kyodo News Agency reported citing US cybersecurity firm.
Of the record-high 770 million scam emails sent globally in May, Proofpoint analysed 240 million containing sender data and found 81.4 per cent of those targeted Japanese speakers, the company said in a recent report.
"Fraudulent emails were easily spotted previously because of unnatural wording, but the advancement of generative AI has helped produce natural sentences, enabling them to break through the language barrier," Proofpoint Japan's Yukimi Sota said.
According to Proofpoint, which says it analyses around a quarter of emails sent globally, the volume of nefarious emails started increasing sharply around the time when Russia began its full-scale invasion of Ukraine in February 2022.
There were 100 million to 200 million such emails sent each month before 2025, but that figure surged to more than 500 million per month this year, the company said.
Many are phishing emails sent from addresses posing as securities firms. They guide recipients to fake websites that are used to steal personal information such as email addresses and passwords, giving hackers the ability to hijack accounts.
If corporate email and security credentials are stolen, it could give attackers access to unauthorised internal communication systems on which further phishing emails can be sent.
According to Sota, the majority of email scams targeting Japan used a specific cybercrime programme that uses the Chinese language. The number of such emails plunged during the Lunar New Year from late January to early February.
"Their unprecedented scale and sophisticated methods raise a possibility of an organised attack led by a foreign government," Sota said, calling on Japanese companies to enhance cybersecurity measures such as adopting multi-factor authentication. - Bernama-Kyodo
