概要: 北朝鮮のIT労働者は、AIツールを使ってID偽装を行い、チャットボットを使ってリアルタイムの面接支援を行い、米国や欧米企業のリモートワークの職を得ていると報じられている。これらの職は機密システムへのアクセスを許可し、意図せずして北朝鮮政権に流入すると思われる収入を生み出しているとの報告がある。
Editor Notes: This incident ID functions as a cluster-level anchor for ongoing reports since 2021 involving suspected North Korean IT workers reportedly using AI tools (such as facial alteration for identity spoofing and chatbot-based interview assistance) to gain employment in Western companies or to impersonate employers in order to deliver malware, exfiltrate credentials, or extract revenue, often as part of reportedly wide-ranging, state-directed efforts to evade sanctions and fund DPRK strategic programs under the auspices of the Reconnaissance General Bureau. The incidents span multiple operations and may involve thousands of actors operating across freelance platforms, third-party facilitators, and front companies. Discrete incident IDs may be developed from this entry as more bounded and independently verifiable cases emerge. While some sources trace this phenomenon to 2014 (with an acceleration during 2020 owing to COVID-19), this incident ID sets 01/01/2021 as its start date to reflect the period in which AI tools were reportedly integrated into infiltration methods. For additional historical context, see: https://en.wikipedia.org/wiki/North_Korean_remote_worker_infiltration_scheme. While there is no confirmed link between Research Center 227 and these specific operations, available reporting indicates that the center functions within a related operational ecosystem. See also Incidents 644 and 1117. U.S. government advisories and indictment press releases are tagged as responses for this incident ID. KnowBe4's reports are also marked with the response tag, but from the standpoint of a company describing itself as a victim of the implicated technology, not its developer.
Alleged: Unknown large language model developers , Unknown deepfake technology developers と OpenAI developed an AI system deployed by Reconnaissance General Bureau , Lazarus Group , Government of North Korea , Department 53 , North Korean threat actors , Famous Chollima , PurpleBravo , WaterPlum , Minh Phuong Ngoc Vong , Sim Hyon-Sop , Kim Sang Man , Christina Chapman , Wagemole , UNC5267 , Void Dokkaebi , Contagious Interview , Gwisin Gang , Matthew Isaac Knoot , Yang Di , Jong Song Hwa , Kim Ryu Song , Ri Kyong Sik , Rim Un Chol , Kim Mu Rim , Cho Chung Pom , Hyon Chol Song , Son Un Chol , Sok Kwang Hyok , Choe Jong Yong , Ko Chung Sok , Kim Ye Won , Jong Kyong Chol と Jang Chol Myong, which harmed Western companies , Companies in the United States , Employers , Cryptocurrency platforms , Developers , Interviewees , macOS users , Recruitment teams , Hiring managers , Human resources staff , Web3 , Blockchain projects , Internal Revenue Service , IRS , Social Security Administration , SSA , Andrew M. , Oleksandr Didenko , Jiho Han , Haoran Xu と Chunji Jin.
関与が疑われるAIシステム: Video interview platforms , Unknown large language models , Unknown deepfake technology , Freelance platforms , Document verification systems , Digital identity verification services , ChatGPT , Laptop farms , BYOD (Bring Your Own Device) , Flashpoint-detected info-stealing malware , Zoom , WebSocket-based C2 , ARP packet signaling , Raspberry Pi Zero , OtterCookie , OtterCookie v3 , OtterCookie v4 , BeaverTail , InvisibleFerret , AgencyHill99 , GitHub , Astrill VPN , FTP exfiltration , Remote admin tools , Job boards , Upwork と remote3
インシデントのステータス
インシデントID
1118
レポート数
37
インシ デント発生日
2021-01-01
エディタ
Daniel Atherton