AltaySec · ATLAS 2026.07
AML.T0082

RAG Credential Harvesting

C · Explainer (cannot be honestly simulated in-browser) AML.TA0013 · Credential Access

Description

Adversaries may attempt to use their access to a large language model (LLM) on the victim's system to collect credentials. Credentials may be stored in internal documents which can inadvertently be ingested into a RAG database, where they can ultimately be retrieved by an AI agent.

Honesty-tier rationale

Credential access is bound to real stores/environments (repos, memory, env vars); it cannot be honestly enacted without infrastructure.

Source

← Back to the matrix