The Cyberattack Refund Has Entered the Apology-Envelope Era
GMO and Park24 plan payments after cyberattacks. Modest goodwill sums and an expected ¥5.5 billion corporate charge put a price on breached personal information.
AI-generated conceptual illustration about corporate compensation after cyberattacks; it is not a documentary image.
Japan’s cyberattack bill has reached the apology-envelope stage.[1]
GMO Internet’s October 24 notice promises ¥5,000, about $32, to all current users of its Onamae.com Desktop Cloud service, regardless of financial loss—not every customer of the wider group. It separately offers compensation for verified realized losses from unauthorized trades linked to the incident, excluding ordinary market losses and missed profits. Park24 plans ¥3,000 payments for roughly 1.46 million Times Car accounts whose identity-document images were transferred outside the company, rather than all 6.62 million compromised accounts. It expects a ¥5.5 billion extraordinary loss for its response and compensation.[1][2][3]
The payments are small enough to feel ceremonial. The corporate charge is large enough to be material. That gap is the story.
The breach bill finds the balance sheet
GMO said it received a demand for money from a party believed to be behind the attack, paired with a threat to expose company and customer information. The company said it would not pay and would work with police and other authorities. Its shares fell almost 16% while Japan’s Topix rose 0.2%, Reuters reported.[1]
Park24’s response contains a more useful operational detail: it says it will stop retaining identity-document images after registration verification. A file no longer retained is one less file available for the next breach. That is a stronger control than another ceremonial apology.[1]
TrendAI counted 723 cybersecurity incidents in Japan through October 20, already above the full-year totals for 2024 and 2025. ESET Japan’s president cautioned that the increase can reflect both more threats and better detection of incidents that previously stayed invisible or unreported. A rising number is evidence of a larger recorded problem, not a clean measurement of attacker growth by itself.[1]
Keep the attack stories separate
Reuters connected the current wave to the critical PHP vulnerabilities that pushed government and corporate sites offline last week. It also stated that GMO did not identify a connection between its breach and those flaws. The temptation to compress every cyber story into one cinematic attack campaign is strong. The evidence here does not support doing that.[1]
The confirmed disclosures are less cinematic: GMO acknowledged customer-information exfiltration; Park24’s payment eligibility turns on stolen identity-document images. The modest planned payments and large expected corporate charge raise a practical question: could collecting and retaining less sensitive information reduce the damage next time?[2][3]
Compensation is better than a free year of credit monitoring and a paragraph about taking privacy seriously. It still does not answer the harder question: what level of liability would change how much sensitive data companies collect, how long they retain it, and how quickly they act on warnings?
What level of liability would make companies keep less sensitive data in the first place?
CORRECTIONS & UPDATES.
Correction · Correction, October 26: clarified which GMO service users and Park24 accounts qualify, distinguished goodwill from verified-loss compensation, and removed unsupported statements about retention motives.
FOLLOW THE SOURCES.
Checked Oct 26, 2026. Numbered references point to the evidence behind factual passages. Commentary is our own.
Two critical PHP flaws pushed banks, trains, tax offices and Japan's mobile emergency-alert system into maintenance mode. The web stack finally sent a very physical invoice.
CrowdStrike found exposed AI-tool records tied to the South Korean bank campaign, including a résumé request. The operator’s identity remains unconfirmed.