AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Japan internet giant GMO says hackers posted stolen customer data

Collected Oct 27, 2026

GMO Internet, the Japanese internet and IT services group, said on October 27 that a suspected attacker had posted information that appeared to be linked to its customers and its business, following unauthorized access to its Onamae Desktop Cloud service. The service, which provides virtualized desktops to financial traders, was shut down on October 21 once the intrusion was discovered. GMO says it is still checking whether the published material is genuine, so the scale of the exposure is not yet established. Its spokesperson did not respond to a request for comment.

GMO also said it believes as many as 27,000 servers may be affected. Customers have been advised to reset their passwords after the company received reports of unauthorized operations carried out on customer trading accounts. On October 24 the company said it would compensate customers and offer an alternative service while seeking help from government agencies, and on October 25 and 26 it began collecting devices from customers to check for malware and better assess the damage. The incident was reported to Japanese police on October 22.

The demand that arrived alongside the breach is unusually large: GMO received an email from the hackers asking for roughly 8 billion yen, about $50 million. The company has not said publicly whether it engaged with that demand. In practice, extortion demands of this size against a listed operator tend to become a regulatory and reputational matter as much as a security one, though the source gives no indication of how GMO is treating the request beyond reporting it.

The breach sits close to separate findings from GMO's own cybersecurity unit, which about a week earlier reported evidence that a Chinese AI-enabled penetration testing tool called Artex had been used in attacks on companies across Japan. GMO has said it has not publicly established any connection between Artex and the Onamae intrusion. CrowdStrike reported earlier this month that Artex was also used against financial institutions in South Korea in September, with those attacks largely exploiting weak security controls such as missing multi-factor authentication. According to Reuters, the tool's Chinese developer has stopped public updates and made it closed-source after it was linked to the bank attacks.

The regulatory backdrop is active. A spokesperson for Japan's Financial Services Agency told TechCrunch that the regulator issued guidance in May urging financial institutions to prepare for a rise in cybersecurity vulnerabilities, calling for stronger baseline defenses in vulnerability management, defense-in-depth and incident preparedness, while encouraging the use of AI and other security tools. Onamae's customer base of traders makes that guidance directly relevant here: virtual desktop services concentrate session credentials, trading access and sometimes client funds infrastructure in one place, so a single intrusion can expose both identity data and the ability to act inside accounts.

That combination is what makes the reported unauthorized operations on trading accounts significant. Password resets address credential theft, but they do little if an attacker held an active session or planted software on customer machines, which is a plausible reading of why GMO started collecting devices on October 25 and 26. This suggests the company is treating endpoint compromise, not just credential exposure, as part of the incident. The device collection is also a practical way to establish whether the posted data is real and how far it reaches.

The 27,000-server figure is worth reading carefully. It describes infrastructure that may be affected, not confirmed compromised customer accounts, and GMO has not broken the number down. For a desktop-as-a-service product, servers and customer desktops do not map one-to-one, so the count is best treated as an upper bound on scope until verification is complete.

A likely trade-off for GMO is speed versus disclosure. Publishing details before the leaked data is authenticated risks amplifying false claims, while staying quiet lets affected traders keep using credentials that may already be in hostile hands. The company's public posture, resets plus device collection plus compensation, leans toward containment first.

Why it matters: Teams running hosted desktop or remote-access products for financial users should expect scrutiny of session handling, MFA coverage and endpoint integrity, since those are the controls the reported attacks elsewhere bypassed. For GMO's customers, the immediate actions are password resets and cooperating with device checks; for product teams more broadly, this is a reminder that a virtual desktop is an account-access system as much as a hosting service. Whether Artex played any role in this breach remains unestablished, and it is my inference, not a reported fact, that investigators will look at it alongside conventional causes.

Read at TechCrunch · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Japan's GMO Internet is verifying alleged customer data posted by suspected hackers following a cloud service breach.