AML.T0110.001 · Implementation

閱讀 MITRE ATLAS 中 AML.T0110.001「Implementation」相關的 1 個 AI 資安案例,查看事件證據、適用範圍與 AIDEFEND 防禦分析。

風險分類說明

Adversaries may poison the executable implementation of an AI agent tool so that normal tool invocation produces unauthorized behavior or hidden side effects. The tool may continue to provide its represented functionality while also accessing additional data, modifying requests, changing recipients or destinations, executing unauthorized commands, weakening security controls, or performing covert data exfiltration. Implementation poisoning does not require the model to interpret or follow malicious instructions. The adversarial effect is produced by executable logic when the agent or user invokes the tool. For example, a poisoned email tool may send the requested message while silently adding an adversary-controlled blind-copy recipient. A file-processing tool may return the requested result while also transmitting the source file to an external service. Implementation poisoning may be introduced before publication, through compromise of a tool's source repository or build process, or through a malicious update after users have adopted a benign version. Installed copies may continue to exhibit the poisoned behavior even after the malicious package or remote listing is removed[[koi-postmark]].

框架官方來源 ↗ · 開啟互動探索 ↗

Incident ·

被投毒的 Postmark MCP server 外洩郵件內容

一名惡意行為者透過供應鏈攻擊,成功從 Postmark MCP server 的使用者身上外洩郵件內容。Postmark 是一項郵件遞送服務,讓組織可以透過 API 寄送行銷與交易郵件;而 Postmark MCP server 則讓使用者能透過 AI agent 操作 Postmark。 這名惡意行為者假冒 Postmark,在 npm 上註冊了 postmark-mcp 這個套件名稱。一開始,他發布的是這個 MCP server 的正常版本。等到這個套件變得熱門、每週下載量超過 1,000 次之後,他就翻臉不認人(rugpull),上傳了一個惡意版本。這個惡意版本會把他自己的電子郵件位址,加進這個 MCP 工具所寄出每一封郵件的 BCC 欄位。凡是升級到這個版本、又繼續使用該工具的使用者,所有郵件都會被外洩到他手上。

沿用 SecureFlow 來源步驟的 ATLAS 標註;不是 MITRE 對跨框架對應的背書。

SecureFlow · 閱讀完整案例 ↗