Daishin Securities
Thirty eight thousand is not a headline. For a brokerage customer list, it is a target list.
What happened?
Daishin Securities, a South Korean brokerage, detected an intrusion on 17 October 2026. Personal and credit information belonging to 38,490 customers was leaked.
The incident sits inside a broader period of pressure on South Korean financial and crypto platforms, including a reported leak at the exchange Bithumb caused by a compromise of an internal mobile system in the same week. Nothing in the available reporting links the two.
What data was actually inside?
Names, email addresses and mobile phone numbers, described in the reporting as personal and credit information.
On its own this looks like the mildest combination in this series. No Social Security equivalent, no account balances, no card numbers. The significance is not in the fields. It is in the list the fields came from.
Who gets hurt and how?
Brokerage customers, which is a population selected for having money and for being accustomed to receiving urgent messages about it.
A verified name, email and mobile number from a securities firm is worth more per record than the same three fields from a retailer, because the pretext writes itself. A message about a margin call, a settlement failure or a suspicious trade arrives on the correct phone number, addressed by the correct name, referencing the correct broker. The recipient has an account there. Every verification signal a cautious person would check comes back clean, and the only false element is the link.
What did they think they were doing right?
The sensitive material was protected. Balances, holdings, account credentials and settlement details are the assets a brokerage defends hardest, and nothing in the reporting suggests those were taken.
That is the validation and also the trap. Defending the crown jewels and treating the contact table as low value is a defensible risk ranking right up until you notice that the contact table is what makes the phishing work. The attacker does not need the balance. The attacker needs to be believed.
What did they not know about their own data?
That a contact list held by a securities firm is not a contact list. It is a pre qualified list of people with investment accounts, and the institutional affiliation is part of the payload.
Data classification schemes almost always rank fields in isolation. Name, low. Email, low. Mobile number, low. The scheme has no way to express that these three fields acquire a different value when the source is a brokerage, because sensitivity is being assessed per column rather than per context.
That is the gap here. Not an unknown system and not an unknown field, but a known field whose risk was computed without reference to where it was sitting.
If your business runs on databases, you probably have similar records, customer data, credentials, financial information. Do you know what's actually in yours?
Not sure where to start? How to find sensitive data across your estate
What does attribution look like the morning after?
No threat actor has been named in the available reporting.
South Korea's notification regime under the Personal Information Protection Act requires notice to affected individuals and to the authorities without delay, and 38,490 is a tractable number to notify. The harder obligation is the warning, because the useful message to these customers is not that their data was taken. It is that the next convincing message about their account is more likely than usual to be fraudulent, and that advice has a short shelf life.
What would have changed the outcome?
Classifying by context rather than by column. The same three fields in a marketing list and in a brokerage customer table carry different risk, and a scheme that cannot say so will under protect the second one every time.
The organisations that handle this well know which of their tables are target lists before somebody else works it out. The NSE Insurance entry in this series covers the same dynamic in a neighbouring industry.
Daishin Securities found out the hard way.
Your team could spend the next 6 months rebuilding systems, notifying customers, and answering legal questions. Or you could spend 24 hours finding out what's actually at risk.
Want to understand the process first? Read our guide to sensitive data discovery .