
AI-generated summary
Following Shinhan Bank's announcement the previous day that the information of approximately 25,000 customers had been leaked due to a new type of hacking using an AI agent, a similar hacking incident occurred at KB Kookmin Bank.
Financial authorities are discussing response measures... Consider spreading additional damage
Blocks after recognizing abnormal external access... Not related to customer financial transactions
(Seoul = Yonhap News) Reporters Han Ji-hoon, Bae Young-kyung, Park Soo-hyun, and Kang Soo-ryun = Following Shinhan Bank, an incident in which customer information was leaked due to hacking occurred at KB Kookmin Bank.
According to financial authorities and the financial sector on the 2nd, KB Kookmin Bank recently confirmed that about 100 pieces of customer information were leaked due to an external intrusion.
It is reported that the bank blocked the server and access path immediately after becoming aware of the possibility that information was leaked through abnormal external access.
The system through which customer information was leaked was known to be a mobile work support system for employees. The bank explains that it has nothing to do with customer financial transactions such as internet banking and mobile banking.
The Financial Services Commission and the Financial Supervisory Service are said to have received a report on the accident from KB Kookmin Bank and are discussing response plans.
Shinhan Bank announced the previous day that the information of about 25,000 customers had been leaked by a new type of hacking using an artificial intelligence (AI) agent.
Korea's two largest commercial banks, KB Kookmin Bank and Shinhan Bank, have been exposed to the risk of AI hacking.
Authorities are said to be conducting an on-site investigation, keeping in mind the possibility of further spread of damage.
AI outlook — possibilities, not facts
Financial authorities will strengthen cybersecurity inspections for KB Kookmin Bank and Shinhan Bank and propose additional security guidelines.
Likely · Within weeks

In relation to the personal information leakage of about 25,000 Shinhan Bank customers, traces of 'ARTEX AI', a Chinese-based AI autonomous penetration testing console, were found on the suspected attack server, but it was not confirmed whether it was actually used, the security industry said.

As accidents caused by AI increase, there is a growing consensus that AI development companies should be held legally responsible, but it is pointed out that it is difficult to apply the existing legal system. The New York Times introduced related public opinion and legal academic opinions, and reported on cases where AI models such as Open AI and Antropic caused security incidents and the process of discussing responsibility.

Following Japan, Naver announced that it will strengthen overseas travel information search results in 45 major Asian regions, including Vietnam, Thailand, Indonesia, the Philippines, Taiwan, Hong Kong, and Macau, and provide weather, exchange rates, visas, entry guidance, safety notices, and AI-based popular restaurant and attraction recommendation services.

Boston Dynamics, an affiliate of Hyundai Motor Group, announced that it had equipped a humanoid robot Atlas with a next-generation robot hand with 13 degrees of freedom and demonstrated precision manipulation by applying tactile sensing technology. The new hand is the size of a human hand, has industrial durability, and is improving adaptability to real environments through a simtureal method.

In relation to the Shinhan Bank customer information leak incident, traces of 'ARTEX AI', a Chinese-based open source AI penetration testing tool, were confirmed on the server believed to have been used in the attack, but whether it was actually used for the attack has not been officially confirmed, a security expert said.

LG U+ announced that it will work with OptAI, a company specializing in AI model optimization, to develop token optimization technology that reduces GPU and power usage and increases processing efficiency. The two companies have secured technology to increase the amount of tokens that can be processed on the same GPU by up to four times the existing size by expanding the existing small language model lightweight technology to the server GPU environment, and jointly held a seminar on the 1st to discuss ways to apply the technology.