Japanese companies have faced a growing wave of cyberattacks in recent weeks, exposing millions of customers’ records as analysts warn that artificial intelligence is helping hackers exploit long-standing security weaknesses and overcome language barriers that once offered some protection.

More than 500 breaches of corporate systems were reported in the year to this month, surpassing the total for all of 2025, the Yomiuri newspaper reported last week.

On October 7, a ransomware attack on IDC Frontier – a Japanese cloud hosting and data centre services provider – paralysed systems used by 495 companies and local authorities in the country.

Those affected included railway operator JR East, which confirmed that more than 6 million customer accounts had been compromised. Used book retailer BookOff also reported that data belonging to 6.43 million members had leaked.

They were far from alone. Online travel firm Skyticket also said this month that a cyberattack on its servers had resulted in 14.6 million users’ data being accessed, while in September, vehicle rental service Times Car reported that 6.6 million members’ accounts had been exposed – potentially compromising the sensitive personal information contained on driving licences, such as residential addresses.

“This is the result of an accumulation of systems that lack adequate security measures,” said Yutaka Sejiyama, deputy director of the Yokohama-based cybersecurity firm Macnica’s Security Research Centre, noting that Japanese companies had long failed to take adequate steps to protect their systems and customers’ private data.

Many had concentrated their efforts on vulnerabilities that could lead to ransomware attacks at the expense of issues deemed lower risk, he said.

It appears that attackers are exploiting vulnerabilities that had gone unnoticed for yearsYutaka Sejiyama, cybersecurity researcher“But security measures are an ongoing process with no definitive end point,” Sejiyama said. “Now, it appears that attackers are exploiting vulnerabilities that had gone unnoticed for years.”He believes recent advances in AI are partly to blame for the latest wave of attacks, though he concedes that “no definitive proof” of this exists as yet.

Still, “the methods employed and the scale of the damage suggest a level of sophistication beyond human capability alone”, he said, further noting that “the malicious use of AI has become commonplace in other cyberattacks”.

Successfully probing a company’s cyberdefences might once have taken a human hacker weeks or months, but observers suspect AI has now hyper-accelerated this process.

Foreign hackers had largely ignored Japan in the past because its language made many systems hard to access and navigate, said Martin Schulz, chief policy economist at Fujitsu’s Global Market Intelligence Unit who analyses AI integration and distributed digital ecosystems.

“Japanese was effectively a barrier,” Schulz said, pointing to the fact that bad actors predominantly used English-language hacking tools.

“But in the last year, AI has progressed brilliantly and they have got around that barrier and can target Japanese organisations now.”Once inside a system, hackers could take advantage of close integration between Japanese companies and their partners, both foreign and domestic, to unleash a “tsunami” of attacks, Schulz said.

Many firms’ defences were unprepared for the onslaught. “Their firewalls have been no match for the AI agents that are now knocking them down at will,” he said.

These breaches “did not come out of nowhere”, Schulz added, pointing to the cyberattack on Jaguar Land Rover that began in August last year, forcing the carmaker to shut down its IT and automated production networks across the UK for five weeks.

Although there was no suggestion that AI agents were used in that attack, the effects were still devastating. Full production did not resume until early 2026. The breach cost the company some £196 million (US$259.5 million) and the wider hit to the UK economy was put at £1.9 billion.

“Because Japanese companies have so much more integration, I believe there is a very real risk that we will see a major breakdown in networks here,” Schulz said.

South Korean firms have faced similar AI-driven attacks recently, including one that hit nine major financial institutions, with cybercriminals successfully compromising customer-facing web services and leaking sensitive data such as names, email addresses and account numbers.

Schulz said he expected other economies in the region to take note as the tools that had allowed hackers to overcome language barriers in Japan would also dismantle other defences that once kept them away. Ordinary consumers can do little to protect themselves, according to Sejiyama.

“While carefully selecting the services and websites where you personal information can be a form of precaution, there are limits to this approach given how deeply IT has permeated society,” he said.

Individuals should watch out for fraud and phishing emails, he added, and concentrate their efforts on limiting the damage after a breach had occurred.

Members of the public had reported an increasing number of suspicious calls from people posing as police officers, he said, as well as emails from contacts they did not recognise. “We recommend remaining vigilant against such scams,” Sejiyama said.