South Korean banks

How they got in, as reported. Automated AI tools that work through many organizations at once.

← All cases
Date
10 Oct 2026
Place
South Korea
Who
South Korean banks · Financial Services
The way in
Automated attacks run by AI agents
Source
BleepingComputer
01 · What was reported

What was reported.

A threat actor targeted multiple South Korean banks including Shinhan Bank, KB Kookmin Bank, and Hana Bank using the ARTEX AI penetration testing suite and Claude agents. The campaign exposed clients' personal and credit card data and caused system outages in some instances.

The way in the source reports: automated AI tools that work through many organizations at once.

A Chinese-speaking hacker launched cyberattacks that shook the South Korean financial sector earlier this month, using the ARTEX AI penetration testing suite and Claude agents.BleepingComputer, added to these pages on 13 Oct 2026. The source
02 · The way in

What automated attacks run by AI agents are.

Automated tools run by AI agents work through many organisations at once. They try the same weakness on each one that answers.

They repeat what a person would do, faster: they read public records, match versions to known flaws and try passwords.

A reading of public records sees what such a tool sees first: the services, versions and records a firm shows to anyone.

It cannot see the tool or whoever runs it. It shows what is open to it.

This part is true of every case of this kind. It says nothing more about South Korean banks than the source does.

03 · What we check for it

What we check for it.

The doors those tools look for first, read externally.

04 · If this happens at your firm

If this happens at your firm.

  • Turn automatic updates on for every computer, phone, website and server.What changes: updates install by themselves, now and then with a restart. It stays on. Choose the hour they run in the update settings and pause one there if it breaks something.
  • Leave nothing that runs your systems reachable from the internet: admin pages, remote desktop and file shares sit behind a private connection or are switched off.What changes: staff connect through the private connection before they reach those pages. It stays on. A page can be opened to one address when it is needed and closed again the same way.
  • Turn on a second sign-in step for every account, email first.What changes: a code or a tap on the phone at each new sign-in. It stays on. A backup code or a second device covers a lost phone and the method is changed in the account's security settings.
The attacks of 2026 and the habit that stops each →
For wealth and advisory firms

What we read for wealth and advisory firms.

  • The records that tell inboxes whether to refuse email sent as you (DMARC, SPF and DKIM), read externally.
  • Web addresses an attacker could register to look like yours, each checked: is it registered, is it set up for mail.
  • Breach records for addresses at your own domain, read externally; nothing inside is touched.
  • The mail records of every other domain your firm owns, read externally.
  • The certificate your website answers with, seen externally: its name, its dates and who issued it.
  • Read with the assessment. What AI assistants answer about your firm. Assistants build their answers from pages they crawl. A look-alike address an attacker could register can be built to be crawled: its robots file invites the AI crawlers in and its page claims your name. When a client asks an assistant for your phone number, they can get the copy's. In 2025, asked for the sign-in pages of 50 brands, an AI assistant gave addresses the brand did not control one time in three. The check is built to ask six assistants: Claude, ChatGPT search, Gemini, Perplexity, Grok and Copilot, through Bing. Monitoring asks again every week and tells you when a copy's phone or address is in an answer. Source: Netcraft research, read on 4 November 2026.

The rule, from its source: What does each US state's data breach law ask a business to do?

Check my firm

See your own firm the way an attacker reads it.

We read your domain from public records and show each way in with a check anyone can re-run.

05 · Source

Source.

Source: BleepingComputer ↗

This page repeats what the source reported, with its date. It says nothing about any firm that is not named in it.