Study Finds Public Container Images Attract Rapid Automated Credential Abuse
Researchers at Stony Brook University reported that automated scanners searched public container images and extracted embedded cloud credentials. In a four-month honeypot experiment spanning eight registries and 300 images, the team recorded more than 75,000 image pulls and over 140,000 AWS test-key API calls. Some OpenAI and Anthropic test keys were also used against AI model APIs. The fastest AWS test-key call occurred 7.86 seconds after publication. Based on key-use records and container execution status, the researchers inferred that static analysis allowed the scanners to retrieve keys without running the containers.
The experiment demonstrates that credentials left in public images can be discovered and used rapidly, with AWS test keys generating API calls as soon as 7.86 seconds after publication. The researchers advised registry operators to check images for secrets before publication and companies to scan them in CI/CD, then revoke or rotate exposed credentials rather than only deleting the public image.
文/王珮瑤|2026-11-20 發表
美國石溪大學(Stony Brook University)研究人員近期發布研究,發現自動化掃描工具持續搜尋Docker Hub等公開儲存庫中的容器映像檔,並擷取其中的雲端服務金鑰。研究團隊透過蜜罐實驗發現,AWS測試金鑰最快在映像檔發布後7.86秒,就被用來呼叫AWS API,顯示企業若不慎將金鑰留在公開映像檔中,金鑰可能很快就遭到未經授權的使用。
這項研究由Billy Tsouvalas與Nick Nikiforakis共同完成,並於2026年11月舉行的資安學術研討會ACM CCS 2026發表。研究團隊歷時4個月,分兩階段在8個容器映像檔儲存庫部署300個蜜罐映像檔,累計記錄超過7.5萬次映像檔拉取活動,以及超過14萬次使用AWS測試金鑰呼叫API的活動。這些API呼叫包括確認金鑰是否有效、查詢AWS帳號中的雲端資源,以及嘗試呼叫可能消耗雲端資源的API。
第一階段研究團隊在Docker Hub發布100個蜜罐映像檔,每個映像檔均含有一組AWS測試金鑰。這100組金鑰全部在映像檔發布後15分鐘內遭到使用,其中56組在1分鐘內就出現API呼叫。此外,部分OpenAI及Anthropic測試金鑰也被用來呼叫AI模型API,耗盡研究團隊預先儲值的額度。
#Stony Brook University#Docker Hub#AWS#OpenAI#Anthropic#Harbor#ACM CCS 2026