Threat Intelligence & Research

Threat Intelligence & Research

We track adversary infrastructure targeting Japan (phishing and look-alike sites, malware delivery infrastructure), DPRK IT-worker networks, and foreign influence operations, combining public reporting with our own observation records. Technical findings that can be released are published under the name of Sola.com Co., Ltd. as research articles and advisories.

Observed facts are separated from assessments, and indicators of compromise are defanged. Tools and services used in our investigations are referred to generically rather than by product name.

Services

What we deliver

01 / Infrastructure Tracking

Infrastructure tracking

We analyze links between domains, IPs, certificates and hosting to follow how attack infrastructure spreads and how its operation changes.

02 / IOC & Detection

IOCs and detection logic

We turn what we track into indicators of compromise and detection logic, delivered in a form monitoring and response teams can use.

03 / Reporting

Reporting

Initial reports, follow-ups and continuous-monitoring reports for public-sector and enterprise clients.

DPRK Threat Research

Research on North Korean cyber threats

Our Threat Intelligence Section analyses North Korea-linked threat groups and the activities of DPRK IT workers, cross-checking public reporting against our own observation records, and publishes what can be released. The articles are in Japanese; English summaries follow.

Observation period 2026.01–03 · Threat Research

Fake online meeting invitations: how DPRK-linked actors abuse recruitment and business meetings

DPRK-linked actors approach their targets as recruiters or business contacts, invite them to an online meeting and send them to a site that imitates a meeting tool. Citing an audio problem or a required update, the site pushes the victim to run a command or install software. In cases attributed to North Korea (Mandiant, JUMPSEC) this led to malware; in similar fake-meeting campaigns without attribution (Netcraft), legitimate remote-management software was handed out instead. Between January and March 2026 we observed at least 38 look-alike domains on a single hosting provider’s infrastructure. On the one site that was live, the “update” was legitimate remote-management software, and installing it notified the operator via Telegram. The operators of these sites have not been identified. The article closes with practical checks for recruiters, engineers, crypto-asset businesses and IT departments.

Read the article (Japanese) →

Announcement of 2026.09.18 · Analysis

Fake interviews and domestic facilitators: the Japan–US–Australia–Germany public attribution of WaterPlum and DPRK IT workers

On 18 September 2026, Japan’s National Police Agency and National Cybersecurity Office, together with the FBI, DC3, Australia’s ASD’s ACSC and Germany’s BND and BfV, published a joint advisory on the WaterPlum group (widely tracked as Contagious Interview) and on DPRK IT workers. The NPA and the FBI assessed that WaterPlum and some DPRK IT workers operate under the 313 General Bureau of the Munitions Industry Department of the Workers’ Party of Korea; Japan’s three earlier attributions concerning North Korea, made alone or with the United States, did not name a specific bureau. The advisory links the two activities through matching IP addresses, and reports an estimated 30,000 or more infected computers in over 100 countries and regions, as well as the first laptop farm identified and dismantled in Japan. We summarise it for non-specialists and list countermeasures for engineers, recruiters and companies that outsource work.

Read the article (Japanese) →

Cases from 2017–2026 · Method

Reading the “fingerprints” of cyberattacks: international attribution methods and DPRK IT worker cases

Attribution rests on several independent lines of evidence rather than a single indicator. We set out seven classes of traces that attackers leave behind (registration, network and servers, site content, code, tactics and behaviour, identity, and money); how frameworks such as the Diamond Model, MITRE ATT&CK, STIX, the Pyramid of Pain, the ODNI guide to cyber attribution and ICD 203 fit together; and five principles for weighing a match, including that sharing a server or IP address is not evidence of a shared operator. We then look at the evidence behind public attributions from WannaCry (2017) to WaterPlum (2026), including the Olympic Destroyer false flag, and apply the approach to DPRK IT worker cases, where identity and money now carry most of the weight (our assessment).

Read the article (Japanese) →

Records from 2022–2026 · Threat Research

Tracking NAVER-impersonating phishing infrastructure, 2022–2026: assessing Kimsuky involvement

We compared our archived scan records with public reporting from 2022 to examine links in domains, IP addresses, certificates and code fragments between NAVER-impersonating phishing sites seen in 2022 and in 2025–2026. The article sets out where the evidence supports the hypothesis of Kimsuky involvement and where the connection remains unconfirmed.

Read the article (Japanese) →

All articles are also collected in the Threat Intelligence Section category (Japanese).

Publications

Articles

DateTypeTitleLang
2026-10-24 Threat Research Tracking NAVER-impersonating phishing infrastructure, 2022–2026: assessing Kimsuky involvement
Our archived scan records compared against 2022 public reporting; domain, IP, certificate and code-fragment links and their limits. IOC appendix and English edition in preparation.
JA
2026-09-22 Landscape Reading Taiwan's cybersecurity ecosystem: ACS, NICS, TWCERT/CC
Taiwan's principal institutions and vendors, with comparisons to South Korea.
JA
2026-08-22 Research Series OVERRIDE 2022
#01 What happened on Japanese-language social media right after the invasion / #02 Not a switch but an accumulation: anti-vaccine and pro-Russia narratives / #03 How the "bioweapons lab" claim became a justification narrative
JA
2026-07-26 Landscape Reading South Korea's cybersecurity ecosystem: KISA, KrCERT/CC, AhnLab, ESRC JA
2026-07-12 Note If an AI execution environment is compromised: risks and protections for Cowork runtimes JA

As follow-ups to this series, we are preparing a 2026 campaign abusing dynamic DNS (infrastructure selection and operational changes) and a comparison of 2022 and 2026.

Media Coverage

In the Press

All media coverage →

AI & Human Judgment

How AI and analysts share the work

We use AI to cross-check public information against our own observation records. Attribution, confidence and publication are decided by our analysts.

AI

What AI does

  • Cross-checks public reporting against our records and proposes candidate related domains, IPs and certificates
  • Drafts and proofreads research articles and reports

Analyst

What analysts decide

  • Attribution of activity and the confidence we place in it
  • Whether IOCs are properly defanged and whether to publish

Policy

Publication Policy

  • Facts and assessments — directly verified facts are stated separately from assessments drawn from them.
  • Defanged indicators — domains and URLs are written with [.]; hashes are given in full.
  • Attribution — co-occurrence or shared hosting alone is not treated as evidence of a common operator; assessments rest on multiple independent lines of evidence.
  • Client information — findings from client engagements are published only after anonymization and with the client's consent.