South Korean brokerages and a crypto exchange have been hit by fresh cyberattacks , weeks after an AI-assisted campaign targeted the cou...
South Korean brokerages and a crypto exchange have been hit by fresh cyberattacks, weeks after an AI-assisted campaign targeted the country's banks. Here is what we know, what experts say, and how customers can protect themselves.
What happened this week
South Korea's financial industry is dealing with another round of cyberattacks, and the pattern is getting harder to ignore. The newest incidents involve the brokerage Daishin Securities, the asset manager Mirae Asset Global Investments, and the crypto exchange Bithumb. All three came to light on Wednesday.
This is not the first wave. Roughly a month earlier, a series of attacks struck South Korean financial institutions, which raised an uncomfortable question: if banks were not the only soft targets, who else was exposed? The answer, it seems, is nearly everyone who handles money.
What links these events is a growing worry about AI-driven hacking across Asia. Investigators are still working out what caused the latest breaches, so some of the story remains unsettled.
Daishin Securities: the confirmed data leak
Daishin is the clearest case so far. The brokerage said hackers exposed the personal information of 38,490 customers. The leaked records include names, email addresses, and mobile phone numbers, with account and phone numbers reported as encrypted. The affected system was a server tied to non-face-to-face account openings for discretionary investment services.
The company says it has notified the customers involved and is working with authorities on the investigation. It also says it will compensate for any losses as required by law.
Those details matter. The server was not a core trading system. It was a supporting service connected to the outside world, which is exactly the kind of weak spot attackers tend to find first.
Mirae Asset and Bithumb: incidents without confirmed data theft
Mirae Asset Global Investments and Bithumb each reported separate cybersecurity incidents on Wednesday. Neither has confirmed that customer data was taken. In Mirae's case, the company said some internal business documents were accessed.
That is a meaningful difference from Daishin, but it is not a reason to relax. An attacker who reaches internal documents has already gotten past at least one layer of defense. Whether that access was a dead end or a stepping stone is what investigators will want to establish.
The September campaign and the AI tool behind it
To understand the current wave, it helps to look at what came before. Security researchers connected the September attacks to a suspected China-based hacker who used an AI penetration testing tool called Artex. Artex is a recently released open-source tool, developed in China, that uses AI agents to carry out penetration testing tasks.
Researchers say the campaign ran from late September into early October 2026 and resulted in stolen data. At one bank, the attacker reportedly got into a loan progress inquiry service used by financial brokers. At another, the target was an employee mobile work-support system. Again, these were side doors, not the front vault.
The attacker has not been tied to any named hacking group. Researchers describe the person as likely a Chinese speaker and financially motivated, though they express only moderate confidence in that assessment. The evidence includes the use of a Chinese-developed tool and prompts written in Chinese.
One detail stands out. The attacker mixed Artex with several large language models and, according to the researchers' findings, even asked an AI assistant where stolen Korean data tends to be sold. That is a worrying picture. AI let one financially motivated operator run several intrusions in a short window, work that once might have taken a team.
Officials are still investigating whether the newest incidents involve the same actor or method. For now, that link has not been established.
North Korean IP addresses appear in the picture
A separate development has added to the alarm. Financial authorities said multiple North Korean IP addresses were identified in attacks that began on the 21st. No breach has been traced back to those addresses, but this is the first time North Korean IP addresses have been spotted in attacks on Korean financial firms.
It is worth being careful here. Spotting an IP address is not the same as proving who carried out an attack, since addresses can be borrowed, routed, or faked. Even so, the finding has unsettled the industry, and it shows how many different kinds of actors are now probing the same targets.
How regulators and the government are responding
South Korea's financial regulator has called for stronger cybersecurity measures after the string of attacks on financial firms and government agencies. Officials in Seoul also said the country plans to run security checks on dozens of IT systems in the private sector.
Those are sensible first steps, but experts argue they do not go far enough on their own. The bigger conversation is about law and governance.
Why experts say the system is vulnerable
Several security professors and industry figures have weighed in, and their arguments overlap in useful ways.
Speed has changed the game
One information security professor at Soonchunhyang University says breaches are now happening across several business sectors at once, at a pace no human team could match. Today's targets are relatively weak servers tied to account openings or business-support networks. The worry is that attackers could use these externally connected systems to push into core internal networks. His recommendations include giving chief information security officers more authority and adopting automated defenses such as attack surface management.
The math favors attackers
An official at a large brokerage described the problem as asymmetric. An attacker only needs to break into one of a hundred firms, so adding staff and budget at a single company cannot stop AI-driven hacking by itself. The major securities firms say they have made abnormal-transaction detection and vulnerability checks routine, but acknowledge that individual defenses are not enough.
Many firms do not know their own networks
A professor at Chung-Ang University who leads a digital authentication association says many financial companies lack a clear picture of their own network layouts and only run surface-level checks after another company suffers an incident. He suggests a hands-on training institute for AI response in the financial sector, along with ongoing red-team and blue-team exercises.
A national AI defense is needed
A Korea University professor proposes that large firms adopt zero-trust frameworks in which AI verifies behavior patterns in real time. For smaller firms, he suggests the government build a security-focused sovereign AI model and an open-source platform, then distribute it cheaply through the cloud.
Data classification matters
Another Korea University professor says AI has sharply increased what attackers, including those in China and North Korea, can do in technology-based crime. He argues companies should classify their data clearly and protect their most important assets first.
The stalled Digital Finance Safety Act
A big part of the criticism points at a law that never passed. The Digital Finance Safety Act, formerly called the Digital Finance Security Act, was pushed by the Financial Services Commission in 2024 and then stalled.
The bill was meant to fix a patchwork problem. Under the current Electronic Financial Transactions Act, security is regulated only through scattered provisions and lower-level supervisory rules. The proposed law would have elevated the Financial Security Institute, now an incorporated association, into a statutory body.Extended oversight to externally linked IT servers replaced checklist-style supervisory rules with legal security responsibility for chief executives and boards
A roadmap from August 2024 on easing network separation rules also cited the law as a key prerequisite for using generative AI in finance.
So why did it stall? Pressing issues crowded it out, including the Hong Kong equity-linked securities losses and the collapse of the e-commerce platforms TMON and WeMakePrice. Then the 21st National Assembly's term ended, and political turmoil followed, including impeachment and an early presidential election. The bill was simply forgotten.
Experts now want the conversation restarted. One cybersecurity professor at Ajou University argues for a framework act on digital finance security with enforcement strong enough to block the launch of new services when security is inadequate. Others want the Financial Security Institute to become a statutory body with investigative powers, similar to the Korea Internet and Security Agency.
A regional pattern, not an isolated problem
South Korea is not alone. A day earlier, Japan reported a cybersecurity incident involving the internet giant GMO Internet, following a series of attacks on its own financial institutions.
Investigators have not established any connection between the Korean and Japanese incidents. Still, the timing points to a wider regional challenge. In many cases, years of weak cybersecurity are catching up with organizations just as AI makes attacking easier and faster.
What customers can do right now
If you use a Korean brokerage, exchange, or any financial service that could be affected, a few habits make a real difference:
Treat unexpected calls, texts, and emails about your account with suspicion. Leaked names, emails, and phone numbers are exactly what scammers use to sound convincing.Never share one-time codes or passwords, even if the person claims to be from the company.
Turn on two-factor authentication wherever it is offered.
Use a unique password for each financial account.
Contact your provider through its official app or website, not through links sent to you.
Watch statements closely for unfamiliar activity and report it quickly.
What comes next
Several questions remain open. Investigators have not announced the root cause of the latest breaches. Mirae and Bithumb have yet to say whether customer data was affected. And no one has confirmed whether the September campaign and the new incidents are connected.
What is already clear is the direction of travel. Attackers are using AI to move faster and hit more targets at once, while the rules protecting financial data are still a patchwork. Until law, governance, and defenses catch up, more incidents like these are likely.
Frequently asked questions
Which South Korean firms were hit by the latest cyberattacks?
Daishin Securities, Mirae Asset Global Investments, and the crypto exchange Bithumb all reported incidents.
How many Daishin customers were affected?
38,490 customers had their information exposed. The records include names, email addresses, and mobile phone numbers, with account and phone numbers reported as encrypted.
Was customer data stolen from Mirae Asset or Bithumb?
Neither company has confirmed that customer data was taken. Mirae said some internal business documents were accessed.
What is Artex?
Artex is a recently released open-source, AI-driven penetration testing tool developed in China. Researchers say a suspected China-based hacker used it in the September attacks on South Korean financial institutions.
Who is behind the attacks?
That is not yet established. Researchers believe the September attacker is likely a Chinese speaker and financially motivated, but no named group has been identified. Authorities have also spotted North Korean IP addresses in attacks that began on the 21st, though no breach has been traced to them.
Are the latest attacks connected to the September attacks?
Officials are still investigating the root cause, and a link has not been confirmed.
How is the government responding?
The financial regulator has called for stronger cybersecurity measures, and the country plans to run security checks on dozens of private-sector IT systems.
What is the Digital Finance Safety Act?
It is a bill the Financial Services Commission pushed in 2024 that would have given the Financial Security Institute statutory status, extended oversight to externally linked servers, and placed legal security responsibility on executives and boards. It stalled and has not passed.
Is this only a South Korean problem?
No. Japan reported an incident involving GMO Internet the day before, though no link between the cases has been established.
What should customers do to stay safe?
Watch for phishing, never share one-time codes, enable two-factor authentication, use unique passwords, and contact providers only through official channels.
Reference
Sedaily - https://en.sedaily.com/finance/2026/10/28/hacked-almost-weekly-koreas-digital-finance-security-bill
Tech crunch - https://techcrunch.com/2026/10/28/south-koreas-financial-industry-hit-by-fresh-wave-of-cyberattacks/
