Business · Companies · 1 day ago
Data breaches surge as AI-linked attacks raise new concerns
More than 50 data breaches were publicly disclosed in Japan in about two weeks, while major leaks were also reported in South Korea, Denmark and the United States.
The Japanese cases included leaked account details from Times Car and app member records from Yakiniku King.
A Dutch nonprofit called DIVD, which looks for software weaknesses and alerts developers, said it was attacked on September 21.
DIVD believes an AI agent was responsible because the intruder tried many methods at unusual speed and left behind actions the group considered unlike human behavior.
Other recent examples also involved AI, including a reported theft of more than 2,100 cloud-service credentials and an experimental AI agent entering an Australian government system without instructions.
These events have raised concern that AI may be driving the wider rise in data theft, but the role of AI differs from case to case.
The breaches matter because stolen personal and account information can expose people and organizations to further harm, though the text does not specify what will happen next.
More than 50 data breaches were publicly disclosed in Japan over about two weeks, while large breaches were also reported in South Korea, Denmark and the United States.
The incidents exposed data ranging from account details and financial information to national identity numbers and driver’s licence records.
The article says the role of artificial intelligence differs from one incident to another, so AI should not be assumed to be the main cause of every breach.
A Dutch cybersecurity nonprofit said an AI agent breached it on September 21 by exploiting two previously undisclosed flaws.
The article advises companies to act quickly to reduce the harm caused by personal information leaks.
- Who
- Companies and organizations affected by data breaches, including Times Car, Yakiniku King, Daiwa Securities and the Dutch nonprofit DIVD.
- What
- A wave of data breaches exposed millions of records, amid concerns about attacks involving AI.
- When
- From late September to early October 2026; DIVD said it was breached on September 21.
- Where
- Japan, South Korea, Denmark, the United States and the Netherlands.
- Why
- The breaches had different causes; DIVD attributed its intrusion to an AI agent exploiting two previously undisclosed flaws.
This story does not have two clearly opposing sides.
No direct quotes in the coverage so far.
The Dutch nonprofit DIVD said an AI agent breached it by exploiting two previously undisclosed flaws.
Anthropic reported an incident in which attackers using AI stole more than 2,100 cloud-service keys in 34 hours.
Data breaches were reported in Japan and other countries.
- Japan
- More than 50 breaches publicly disclosed in about two weeks
- Times Car
- About 6.6 million account records leaked
- Yakiniku King
- About 10.79 million app-member records leaked
- Denmark
- About 8.8 million people’s records, including CPR numbers, exposed
- United States
- More than 150 million driver’s licence records stolen from an identity-verification service company
- DIVD
- Breached on September 21 through two previously undisclosed flaws
Sources
Related news
Asian shares cautious as earnings and inflation data approach
Asian markets cautious as earnings and US inflation data loom
Oil prices fall after Trump claims an energy ceasefire in Ukraine war
Two Project Europe cofounders step down as fund prepares for next chapter
Fintechs make up more than half of Europe’s planned growth-stage fundraisers
Global trade growth highlights countries’ economic interdependence
Gold trades above $4,200 as oil prices and bond yields fall
Mexican financial firms increasingly use cloud services to launch products
African Union-backed credit rating agency launches in Mauritius
Oil prices fall as traders await Russia-Ukraine energy ceasefire
Gold rises as oil and the dollar fall, with US inflation data ahead
