Japan’s Cybersecurity and Institutional Resilience Crisis
An evidence-based snapshot of the incidents disclosed in Japan in 2026, and a living investigation of the accountability model behind them.
- Version
- 0.1
- Evidence up to
- 20 November 2026, Japan time
- Edition
- Free edition
- Length
- 16 chapters, 58 sources, about 41 minutes
Working analytical report. Findings are provisional where investigations remain open.
Ringi Review

Japan’s Cybersecurity and Institutional Resilience Crisis
Version 0.120 November 2026
One intrusion at one supplier. Twenty of the notices that followed. The figures use different units and overlap, so they are never added.
PLAID KARTE
One intrusion, 13 November 2026. Attackers exploited a server vulnerability, stole connection credentials and extracted data.
Confirmed exfiltrationSelect a slip for what its mark means and its source.
Shown in the report’s order, largest stated figure first, not in order of time. The figures use different units and overlap, so they are never added. See the same data as a table.
Contents
- Executive summary
- Scope, definitions and method
- The PLAID KARTE supplier cascade: a systemic dependency case
- Independent financial-sector incidents and customer-data risk (full report)
- Ecommerce and healthcare: basic engineering controls and permanent data (full report)
- Public-sector systems: supply chains, integrity and availability (full report)
- Institutional thesis: the software outsourcing and accountability model
- Cross-case failure taxonomy (full report)
- Consequences for consumers, institutions and society (full report)
- Is Japan uniquely exposed? A careful international comparison (full report)
- Findings ranked by evidentiary strength (full report)
- Action agenda: what competent accountability would require (full report)
- Research agenda and falsification tests (full report)
- Living-report maintenance protocol (full report)
- Appendix A. Source register
- Appendix B. Change log
Executive summary
Japan is confronting a serious test of digital resilience, and the recent cases expose more than a high volume of cyberattacks. The most consequential pattern is a recurring failure to keep sensitive data, application privileges and operational dependencies within accountable boundaries. A single intrusion at
PLAID's KARTE platform produced an extraordinary cascade of corporate notices. Other independent incidents exposed weaknesses in test-data governance, historical payment-card retention, ecommerce application security and government technology operations. Several administrative failures occurred without any attacker at all. S01–S05, S12–S18, S25–S29
Five findings define this snapshot.
- A supplier can become a systemic customer-data dependency. Attackers exploited a
PLAID server vulnerability on 13 November, stole system/database connection credentials and some external-service secrets, and extracted information. This was one supplier-origin incident, not dozens of independent attacks on PLAID clients. Disclosures from financial firms, retailers, telecommunications providers, educational services and others demonstrate how the consequences propagate. Counts in client notices use incompatible units and cannot be added into a national victim total. S01–S09 - Data often outlives its legitimate operational purpose. Former KARTE clients say their data remained after contracts ended;
HAY Japan identifies a 90-day deletion obligation.
MUJI disclosed historical full credit-card numbers from an earlier loyalty scheme, while
CDJapan confirmed extraction of years-old card records. This is a repeatable governance problem: data ownership is not complete without tested deletion and evidence of completion. S06–S08, S14–S16 - Testing and integration environments can be production-risk systems.
Rakuten Securities' possible exposure arose from a test environment holding copied real customer data; a separate municipal migration in Inazawa inadvertently submitted test records to the actual resident-register network, invalidating cards and certificates. These are not the same failure, but both challenge the assumption that testing is inherently lower risk. S12–S13, S28 - The failure mechanism is not exclusively cybersecurity. Some cases concern compromised payment-page code, SQL injection, exposed identity-document images, an exploited analytics component and a shut-down pharmaceutical-information platform. Others concern an account-recovery experience, delayed supplier notification or an ordinary software migration mistake. The combined evidence suggests a broader problem of engineering ownership, lifecycle assurance and management decision-making, not merely inadequate perimeter security. S17–S18, S20–S29
- There is meaningful official support for a structural diagnosis, but causality still has to be proved. Japan's Digital Agency, METI and IPA have described legacy systems, weak internal IT autonomy, entrenched vendor relationships, multilayer subcontracting and vendor lock-in. A Digital Agency review cites legacy systems at 61% of surveyed user companies. That corroborates a problematic industry structure; it does not prove that outsourcing caused each incident, that a handful of integrators dominate every contract, or that Japan ranks first worldwide for cyber exposure. S30–S34
- In use
- Kept after use ended
- Deletion asked or due
- The 2026 event
Point at a case, or select it, for its dates and its sources.
Dark: the stated period of use. Light: the time the data stayed after. A dotted start means the sources give no start date. The red point is the 2026 event. Sources: S06, S07, S08, S14, S15, S16, S25, S26, S27, S47.
The central thesis to investigate: Japan's institutions may have built a fragile dependency model in which software is frequently treated as a procured cost rather than a strategic, internally understood capability. Complex supplier chains and siloed management may weaken ownership of architecture, vulnerabilities, credentials, data retention, incident response and recovery. The most important distinction is between organizations showing evidence of security activity and those proving the effectiveness of controls over time. This thesis is consistent with important cases and official analysis, but remains a causal hypothesis rather than a universally established conclusion.
The strongest immediate recommendation: Regulators, boards and procurement leaders should demand an inspectable inventory of critical vendors and data flows; explicit accountability for security decisions; independently tested secret isolation, test/production separation, payment-page integrity and deletion controls; and auditable time-to-notification and recovery measures. A certificate, security presentation or contract clause is not sufficient proof that the related control actually works.
Key facts at a glance
| Evidence | Interpretation | Critical caveat |
|---|---|---|
| Verified common supplier compromise | Not every client figure is confirmed exfiltration | |
| Serious financial-data confidentiality risk | The datasets' populations and record units require careful treatment; no proof all data was extracted | |
| Retention and storage risk | Possible access, not proven exfiltration; ~345,000 cards reported still valid | |
| Confirmed database theft and separate checkout skimming | Subsets overlap; checkout views are not unique cardholders | |
| Hokkaido: Iwamizawa, Shintoku and Rumoi affected through a common GIS Hokkaido → Rimgate path | Documented multi-client subcontractor blast radius | Potential exposure, not confirmed theft |
| Inazawa: 17 cards and 43 electronic certificates invalidated after misdirected test data | Confirmed integrity/availability consequences without an attacker | Certificate counts can overlap with card holders |
| Digital Agency: 61% of surveyed user firms retain legacy systems | Structural concern acknowledged officially | Survey result, not an estimate for every Japanese firm |
Scope, definitions and method
This is an incident-based investigative snapshot, not an official census of Japanese cybercrime, an audit of the named companies or a ranking of countries. It starts from contemporary notices, technical explanations, investigative reporting and public posts on social media. Where possible, it privileges first-party statements from the compromised organization, affected client, regulator or municipality. A press interview, independent tracker or social-media account is identified separately when it supplies information not present in official disclosures. Links and publication dates are preserved to make revision possible.
Confidence categories used here:
- Confirmed incident
Confirmed incident
A company or authority acknowledges the intrusion, outage or error.
- Confirmed exfiltration
Confirmed exfiltration
The source reports that data left the controlled environment.
- Potentially exposed
Potentially exposed
Data may have been accessible; it is not established that it was taken.
- Operational impact
Operational impact
A service stopped or a credential failed, with or without a data breach.
- Reported by media
Reported by media
Credible reporting, not confirmed by a public primary document.
- First-person post
First-person post
Evidence of a user’s experience, not proof of a technical cause.
- Hypothesis
Hypothesis
A causal explanation that still needs a measurable test.
- Confirmed incident/root: A company or authority acknowledges the underlying intrusion, exfiltration, service outage or operational error.
- Confirmed exfiltration: The source affirmatively reports data left the controlled environment; this does not automatically confirm every field or client maximum.
- Potentially exposed: The affected data may have been accessible or transmitted; investigations have not established that all were taken.
- Operational impact: A service stopped or identity credential failed; this may occur without a personal-data breach.
- Reported by media / interview: Credible reporting attributable to sources, but not necessarily independently corroborated by a public primary document.
- First-person allegation / social reaction: Useful evidence about user experience or trust, not proof of a technical cause or wider prevalence.
- Hypothesis: A causal or institutional explanation needing measurable testing.
Measurement rules: Treat client disclosures from a single supplier attack as one root incident with multiple notifications. Preserve records, accounts, files, page views, customers, land parcels, buildings and unique people as distinct units. Maximum possible exposure is not an observed exfiltration count. Related client populations can overlap. Two data categories within one organization may also describe the same individual. "No confirmed misuse" is a time-bounded statement, not a guarantee that misuse did not happen later.
Timing: This version covers sources obtained through 20 November 2026 in JST, but many ongoing incidents will change. A disclosure dated November 19 may reflect a November 4 intrusion or data from 2019. Disclosure volume does not equal attack volume. The research sample is enriched for widely reported and large incidents, making it unsuitable for estimating the national prevalence of insecurity.
The
PLAID KARTE supplier cascade: a systemic dependency case
Root intrusion: what
PLAID itself confirmed
On 13 November 2026, attackers exploited a vulnerability on
PLAID/KARTE infrastructure, stole credentials used for system and database connections, gained unauthorized access to administrative components and extracted some data. PLAID also reported compromise of at least some integration credentials used with external services and information belonging to customer organizations. It detected unauthorized access at 08:59 that day and subsequently invalidated credentials, shut servers and placed KARTE services into emergency maintenance. Its 16 November notice confirms that some entrusted data, potentially including personal information, left the environment. The precise vulnerable component, attacker path, tenant boundaries and the exact extent of exposure were still being investigated. S01–S02
The security consequences extend along two pathways: confidentiality (data extracted or potentially accessible) and availability/dependency (companies losing access to KARTE functions or pausing their integrations). A company's emergency shutdown in response to the supplier event is not evidence that its own databases were compromised. Similarly, a customer's maximum potentially affected records is not proof that the attacker extracted that maximum.
The
PLAID event should be visualized as one supplier-origin root linked to numerous client disclosures and service effects, not treated as multiple new independent intrusions. Independent incident indexes assembled dozens of company notices as of 19 November; at least one running index listed around 96 related organizations, but its changing taxonomy may include service disruption and indirect clients. That provisional count is not a certified set of distinct breach victims. S03
Selected client disclosures and evidence status
The following table is an inventory of selected, consequential disclosures, not a sum or a complete list of every organization. As the public record evolves, its counts and scopes may change.
| Affected organization/service | Publicly stated figure | Data / risk discussed | Evidence distinction |
|---|---|---|---|
| ~19.11m records | Customer contact and demographic fields | Client-reported leakage; deduplication and fields are not confirmed across notices S03 | |
| Up to ~18.29m accounts | Membership-related information | Maximum potentially affected; not confirmed unique exfiltrations S03 | |
| Up to ~8.58m and additional purchase-history categories | Retail customer and transaction datasets | Figures may overlap and use different grains S03 | |
| ~7m member-related records | Membership data | Supplier cascade, not a new | |
| Up to ~6.97m records | Member information | Reported total may contain duplicates S03 | |
| ~6.5m records | Gender, age, geographic region | ||
| ~5.82m records | Email / hashed user identifiers | Supplier-related disclosure; count is records S03 | |
| Up to ~5.98m members | Retail member information | Potential exposure, not all confirmed stolen S03 | |
| Up to ~4.06m member records | Travel customer information | Potential exposure S03 | |
| Up to 3,264,445 records | Age, birthday, sex, KARTE internal ID | Potential exposure; passwords/card data not sent to KARTE S05 | |
| ~2.38m identifiers or demographic records | Ticketing / demographic information | Different identifiers may not map one-to-one to people S03 | |
| Up to ~1.42m records | Email, demographics, geography and certain subscription dates | Includes some former customers from 2020 onward; no payment or login credentials reportedly supplied S09 | |
| ~948,000 records reported | Age/sex and possibly product-related information | Media-reported scope; no first-party confirmation of each field S03 | |
| ~331,586 records | Names, contact details and possible visit-related records | Some former members; no card/passport data reported S03 | |
| Up to 226,662 people potentially affected | Email, some names/phones; associated customer IDs | Job-seeking worker data reportedly not impacted; ID counts can exceed persons S10 | |
| ~150,069 historical email/ID records | Former marketing data; separate | Contract ended in June 2026; data categories must not be added without deduplication S03 | |
| Up to 9,890 records | Names, emails, member IDs | Former-client data remained after contractual deletion deadline S07 | |
| Up to 29,950 records in company's notice | Employment, skills, recruitment/interactions and contacts | Former-client retention; deletion evidence requested S08 | |
| Count under investigation | Former | Some services ended in 2019 or 2024 S06 |
Client containment counterexample.
Sharp's announcement says it did not synchronize passwords or credit-card information with KARTE. This is a meaningful limit on the supplier's blast radius. A supplier breach is not proof that all customer data held by every affected company leaked; minimum-necessary data transfer can materially reduce the damage. S05
Another useful counterexample. Some financial companies that depended on
PLAID-related services publicly stated that their customers' information had not been confirmed leaked. For instance,
au Jibun Bank's November 19 notice said it used certain QANT/KARTE components but had not received confirmation of customer-information leakage. Such notices belong under operational dependency or precautionary monitoring, not automatically under confirmed exposure. S11
Former clients say their data was still at the supplier long after their contracts ended. This part follows the retention trail, the stolen integration secrets, the security work PLAID had published, and what MUFG was told before the breach.
- Post-contract data retention: the clearest observable governance failure
- Credentials, integration tokens and lateral exposure
- PLAID's security program and the difference between activity and outcomes
- MUFG and bespoke security assurances
- Leadership relationships and independent challenge
In the full reportAbout 4 minutes of reading
Independent financial-sector incidents and customer-data risk
Real customer data in a brokerage’s test environment. Full card numbers kept after two reported requests to delete them. Identity documents held by a contractor. For each case: what is confirmed, and what is only possible.
- Rakuten Securities: high-value financial data in a test environment
- Rumors, fear and access recovery following Rakuten's disclosure
- MUJI / Ryohin Keikaku: historical full card numbers
- Monex Securities and SBI Shinsei Bank: outsourced KYC onboarding
- Older Rakuten unauthorized-trading claims: a separate historical comparison
In the full reportAbout 6 minutes of reading, with the plate “The same line, crossed in both directions”
Ecommerce and healthcare: basic engineering controls and permanent data
A checkout page can leak cards that the shop never stored. Five cases, from SQL injection to passwords stored without encryption, and the control that would have stopped each one.
- Neowing / CDJapan: confirmed extraction plus web skimming
- Rokkatei: active checkout compromise despite no card storage
- SBC Medical / Shonan Beauty: exposed health-service profiles and credential storage
- NewsPicks / Uzabase: third-party analytics software as an attack path
- Sharp's separate 2024 checkout incident
In the full reportAbout 3 minutes of reading
Public-sector systems: supply chains, integrity and availability
One subcontractor, three municipalities, and data from 2019 still on a device in 2026. Then the case with no attacker at all: test records that invalidated real identity cards.
- Hokkaido: one subcontractor incident, at least three municipal clients
- Inazawa: a non-malicious migration error caused genuine identity outages
- MHLW: a recently launched pharmaceutical supply-information system
- More than cyberattacks: government IT as a public-service delivery problem
In the full reportAbout 3 minutes of reading, with the plate “From detection to the people concerned”
Institutional thesis: the software outsourcing and accountability model
What Japan's own agencies have said
The argument that software engineering is sometimes treated as a second-class internal capability is not simply an outsider's cultural critique. In an August 2025 Digital Agency interview, a METI official described Japanese organizations treating IT as a cost, outsourcing systems to vendors rather than maintaining internal knowledge, and entering entrenched user–vendor relationships, layered subcontracting and customized systems. The same modernization review cited legacy systems at 61% of surveyed user companies. METI's May 2025 modernization report warns that legacy environments obstruct digital adoption and competitiveness. IPA likewise points to inherited user–vendor relationships and development practices as obstacles. S30–S32
In 2026, the Digital Agency's procurement reform deliberations addressed vendor lock-in, buyer-side technical capability, interoperability, access for new providers and practical exit options. The Japan Fair Trade Commission launched a November inquiry into local-government IT procurement and vendor-switching issues. These actions indicate that government recognizes the structural questions; they are not evidence that every government vendor is negligent. S33–S34
Proposed mechanism to test
Possible chain: IT viewed primarily as a purchased cost → insufficient internal architecture and software ownership → dependence on established prime integrators → multilayer subcontracting and bespoke systems → fragmentation of responsibility for patching, credentials, data inventories, deletion and incident reporting → slow cross-functional remediation and a larger shared failure surface.
- IT viewed primarily as a purchased cost
- Insufficient internal architecture and software ownership
- Dependence on established prime integrators
- Multilayer subcontracting and bespoke systems
- Fragmented responsibility for patching, credentials, data inventories, deletion and incident reporting
- Slow cross-functional remediation and a larger shared failure surface
The chain fits several observed facts. It does not prove that those failures had the same origin.
This mechanism fits several observable facts: Iwamizawa's data passed through multiple contracted entities;
Monex's onboarding system was maintained by an unnamed third party;
PLAID's client platforms concentrated sensitive marketing data and secrets;
HAY's post-contract data remained at its provider. It does not prove those failures had the same causal origin. The direct technical cause of a SQL injection, for example, may be poor application design regardless of whether a developer was an employee or a contractor.
Vendor market concentration remains an open measurement problem
A strong claim that "only a few system integrators take all the outsourcing" cannot be made solely from anecdotes or recognizable vendor brands. It requires procurement spend data and contract awards, broken down by ministry, municipality, industry, system layer and prime-versus-subcontractor role. A suitable empirical study would measure contractor shares of spending, repeat awards, sole-bid frequency, number of subcontracting hops, software portability and buyer switching costs. The Digital Agency and competition authority provide reasons to investigate; they do not establish a concentration statistic. S33–S34
Organization-wide coordination and
Docomo's network-quality debate
A November 2026 public commentary about NTT
Docomo argued that engineering network quality requires procurement, construction, site negotiations, finance, sales and network teams to pursue a company-level objective; individual departments may optimize their own budgets while overall service quality suffers. This is a management interpretation, not proof of Docomo's private decision history. Independent network measurements show mixed outcomes: Docomo has strengths in coverage and 5G download performance, while competitors lead some reliability and consistency metrics. Docomo also increased planned capital expenditure in November 2026. S51
The useful generalizable question is whether executives can identify and override local incentives when a technology problem demands coordinated decisions. It applies to cyber-risk escalation as an organizational mechanism, but the
Docomo performance dispute is not itself another cyber incident.
Board composition, global perspective and real independent voice
A 2025 Spencer Stuart comparison reports foreign directors occupying approximately 5% of Nikkei 225 board seats and 8% of TOPIX 100 seats. This documents relatively low international representation in the sample, not an absolute absence of foreigners on Japanese boards. Some large corporations, including
Rakuten Group and
MUFG, have foreign directors. Name, photograph and country of origin cannot establish engineering competence or a director's actual influence. S41, S52
One working hypothesis is that board meetings and materials may be overwhelmingly Japanese-language and foreign directors may sometimes lack the information, interpretation or informal access to challenge incumbent management. There is independent support for language being a barrier, and some companies provide translated materials or simultaneous interpretation. These are varying practices, not evidence that every foreign director is symbolic. A test of this hypothesis would measure board-pack delivery times, language and translation quality, attendance and challenge, cybersecurity expertise, dissent and management response, not assume ethnicity predicts judgment. S53
The strongest governance metric is substantive independent technical oversight: whether someone with appropriate expertise can obtain complete information, challenge management, demand remediation and track closure regardless of nationality.
Cross-case failure taxonomy
Nine kinds of failure, the cases that show each one, and what an auditor can test for. One table that turns the incidents into a checklist.
In the full reportAbout 1 minute of reading
Consequences for consumers, institutions and society
An email address is one kind of loss. A home address with declared assets is another. What each kind of exposure can lead to, what is rumour, and what redress should look like.
- Data exposure is not one type of harm
- Public reaction and loss of trust
- User authentication and transaction fraud
- Customer redress cannot end at password resetting
In the full reportAbout 2 minutes of reading
Is Japan uniquely exposed? A careful international comparison
Is Japan worse than other countries, or only more visible? The survey figures, the government’s own warning, and the foreign cases that cut against a simple answer.
In the full reportAbout 1 minute of reading
Findings ranked by evidentiary strength
Every conclusion sorted into three groups: strongly supported, serious questions, and claims the evidence does not support. Read this list before you quote the report.
- Strongly supported
- Supported as serious questions, not proven causal explanations
- Not established by the evidence
In the full reportAbout 1 minute of reading, with the plate “Every finding, by the strength of its evidence”
Action agenda: what competent accountability would require
What a board should demand now, within 90 days and over a year, and what procurement, engineering and regulators should each change. Written as outcomes that can be measured.
- For boards and executive managementThree time frames: what to assign now, what to commission within 90 days, and what to build over a year. With the measures a board dashboard should carry: control outcomes, not activity.
- For procurement and system-integrator governanceThe terms to write into every contract: maximum retention, deletion certificates, test rights, notification deadlines, an exit plan. And the offboarding test that stops a subcontractor keeping data for years.
- For engineering and security teamsFive areas with concrete rules: identity and secrets, data lifecycle, application integrity, release and recovery, monitoring. Written against the failures in this report.
- For regulators and public-sector buyersAccountable owners for outsourced systems, sample audits of deletion, and one identifier per root incident, so that one supplier breach is never counted as eighty attacks.
In the full reportAbout 2 minutes of reading
Research agenda and falsification tests
Twelve open questions, the evidence that would answer each one, and the finding that would prove the thesis wrong.
In the full reportAbout 2 minutes of reading
Living-report maintenance protocol
How this report is kept current: what each revision states, the correction rules, and what version 0.2 will cover.
In the full reportAbout 1 minute of reading
Appendix A. Source register
The source register is organized by topic. Primary denotes a company, municipal, regulator or governmental source; secondary denotes independent reporting or analysis. Some detailed article bodies are restricted; any fact relying on a headline/preview is labeled accordingly in the body.
PLAID / KARTE and clients
- S01Primary: PLAID, 16 Nov 2026. Unauthorized access and leaked information, root event. plaid.co.jp
- S02Primary: PLAID, incident status updates, 13–20 Nov 2026. Response sequence and service effects. plaid.co.jp and status.karte.io
- S03Independent case index and client notices. Broad disclosure index; counts provisional. japanhacked.jp and piyolog.hatenadiary.jp
- S04Primary: MUFG Bank, 18 Nov 2026. Customer demographic records implicated in KARTE incident. bk.mufg.jp
- S05Primary: Sharp, 19 Nov 2026. COCORO STORE maximum 3,264,445 records. corporate.jp.sharp
- S06Primary: Sankei Digital, 18 Nov 2026. Former KARTE services and retained information. sankei-digital.co.jp
- S07Primary: HAY Japan, 18 Nov 2026. Contract ended June 2025; 90-day deletion and data exposure. hay-japan.com
- S08Primary: Offers / overflow, 17 Nov 2026. Former-client data and deletion-proof request. offers.jp
- S09Primary: OPTAGE, 18 Nov 2026. Mineo potential exposure including former subscribers. optage.co.jp
- S10Primary: Timee, 19 Nov 2026. B2B customer-contact information potentially affected. corp.timee.co.jp
- S11Primary: au Jibun Bank, 19 Nov 2026. PLAID dependency without confirmed customer-info leakage. jibunbank.co.jp
Financial-sector incidents
- S12Primary: Rakuten Securities, 17 Nov 2026. Test-environment exposure and reported account categories. rakuten-sec.co.jp
- S13Secondary: ITmedia, 18 Nov 2026. Rakuten's explanation of real production data in testing. itmedia.co.jp
- S14Primary: Ryohin Keikaku, 4 Nov 2026. MUJI data incident second notice. ryohin-keikaku.jp
- S15Primary: Credit Saison, 4 Nov 2026. Historical MUJI Card numbers and issuer exposure notice. saisoncard.co.jp
- S20Primary: Monex Securities, 13 Nov 2026. Outsourced onboarding service and KYC files. prtimes.jp
- S21Primary: SBI Shinsei Bank, 18 Nov 2026. Monex financial-intermediation partner notice. sbishinseibank.co.jp
- S43Primary: Rakuten Securities, 19 Nov 2026. Receipt of FSA report-demand order. rakuten-sec.co.jp
- S46Primary: Rakuten Securities, 18 Nov 2026. Shibuya physical-store opening postponed, reason unspecified. rakuten-sec.co.jp
- S47Secondary: Nikkei XTECH, 18 Nov 2026. Credit Saison reportedly requested MUJI data deletion twice. Accessible headline/syndicated synopsis; full text restricted. xtech.nikkei.com
- S48Secondary: Bloomberg / Courrier Japon, 2025. Rakuten unauthorized-trading customer-remedy dispute; historical comparison only. courrier.jp
Application, ecommerce and healthcare cases
- S16Primary: Neowing Japanese report, 18 Nov 2026. SQL injection, confirmed extraction and skimming. neowing.co.jp
- S17Primary: CDJapan English report, 18 Nov 2026. International site subset of Neowing incident. cdjapan.co.jp
- S18Primary: Rokkatei, 18 Nov 2026. Compromised online-shop checkout page. rokkatei.co.jp
- S22Primary: SBC Medical Group, 19 Nov 2026. Database access and external booking service. prtimes.jp
- S23Primary: SBC Medical Group corporate statement, 19 Nov 2026; secondary coverage. ir.sbc-holdings.com and sbbit.jp
- S24Primary: Uzabase / NewsPicks, 18 Nov 2026. Metabase vulnerability and forensic investigation. corp.newspicks.com
- S49PCI Security Standards Council. Payment-page e-skimming guidance. blog.pcisecuritystandards.org
- S50Primary: Sharp, 30 Oct 2024. Separate earlier COCORO STORE incident. corporate.jp.sharp
Public-sector systems
- S25Primary: Iwamizawa City, 18 Nov 2026. GIS Hokkaido/Rimgate potential exposure. city.iwamizawa.hokkaido.jp
- S26Primary: Shintoku Town, 18 Nov 2026. 2019 commissioned evacuation-system data and timeline. shintoku-town.jp
- S27Kyodo syndication, 19 Nov 2026. Rumoi and Shintoku cross-client reporting. 47news.jp and oita-press.co.jp
- S28Primary: Inazawa City, 13 Oct 2026. Test-data transmission and My Number impact. city.inazawa.aichi.jp
- S29Primary: MHLW, 18 Nov 2026. Pharmaceutical supply system intrusion; official launch context. mhlw.go.jp and mhlw.go.jp
Structural, governance and security-assurance evidence
- S30Primary: Digital Agency interview, 27 Aug 2025. User-vendor dependency, IT as cost and multilayer subcontracting. digital-agency-news.digital.go.jp
- S31Primary: METI, 28 May 2025. Legacy Systems Modernization Committee report. meti.go.jp
- S32Primary: IPA. Promotion of software modernization, legacy development practices. ipa.go.jp
- S33Primary: Japan Digital Agency, FY2026. Procurement reform and vendor lock-in. digital.go.jp
- S34Primary: Japan Fair Trade Commission, 4 Nov 2026. Local-government IT procurement inquiry. cms03.jftc.go.jp
- S35Primary: PLAID engineering, 21 Dec 2022. API tokens held in databases and log secret detection. tech.plaid.co.jp
- S36Primary: PLAID engineering, 30 Jun 2020. STRIDE threat analysis. tech.plaid.co.jp
- S37Primary: PLAID engineering, 29 Nov 2023. Internal security-review practices. tech.plaid.co.jp
- S38ISMS registration. PLAID ISO 27001/27017 scope and validity. isms.jp and isms.jp
- S39Primary: PLAID, 17 Jul 2025. Audit logs and regional service plan. plaid.co.jp
- S40Interview: Diamond Harvard Business Review, 22 Jan 2024. MUFG/PLAID bespoke-security relationship. dhbr.diamond.jp
- S41Primary: PLAID. Leadership biographies and corporate governance. plaid.co.jp and plaid.co.jp
- S42Primary: PLAID, 2025. Kurahashi/Takayanagi professional-network history. blog.plaid.co.jp
- S51Secondary: independent mobile measurements and company capital-investment reporting, October–November 2026. prtimes.jp and itmedia.co.jp
- S52Spencer Stuart, 2025 Japan Board Index. Foreign-director representation. spencerstuart.com
- S53JPX, Egon Zehnder, OECD. Board language and governance access. jpx.co.jp ; egonzehnder.com ; oecd.org
Consequences, comparative context and government warning
- S44Public posts, November 2026. First-person accounts and commentary in public posts on social media; not authenticated as corporate facts. Official Rakuten incident notice governs technical assertions. rakuten-sec.co.jp
- S45Primary: Japan Personal Information Protection Commission / National Police Agency. Illicit personal-information lists, organized crime and practical home safety. ppc.go.jp and npa.go.jp
- S54NordPass 2026 password list and independent academic comparison. nordpass.com and link.springer.com
- S55Secondary: TVQ, November 2026. Consumer PayPay/LINE alleged account takeover and loss. tvq.co.jp
- S56Reuters, 12 Nov 2026. Survey of 214 Japanese company respondents on cyberattacks and supplier exposure. reuters.com
- S57Independent Japan Cybersecurity Tracker, weekly data. Organizational notices are not independent incidents. mkultraman.com
- S58Primary: National Cybersecurity Office, 5 Nov 2026. Fundamental security controls advisory. cyber.go.jp
- S59Reuters, 19 Nov 2026. Third-party incidents involving JPMorgan/Visa/Mastercard. reuters.com
Appendix B. Change log
| Version | Date | Change |
|---|---|---|
| 0.1 | 20 Nov 2026 | First edition. Evidence to 20 November 2026 from publicly accessible primary statements, selected independent reporting and explicitly attributed public reactions, including the 19 Nov FSA report-demand order and the 5 Nov national cybersecurity advisory. |
| 0.1, correction | 20 Nov 2026 | Company name corrected: the clinic booking service called “Triview” in the first edition is |
This is an evolving investigation, not a final finding of national or individual fault.
© 2026 Pulse Browser, Inc. All rights reserved. You may quote this report with its title and date. You may not copy it, share the full report or put it online.
Japan’s Cybersecurity and Institutional Resilience CrisisFull report and future versions, ¥5,000