AML.CS0035
Data Exfiltration from Slack AI via Indirect Prompt Injection
- Type
- Exercise
- When
- 2024-08-20
- Actor, as ATLAS names it
- Not stated
- Target
- Slack AI
- ATLAS attributes it to PromptArmor (security evaluation).
Techniques, in order
- 1Retrieval Content CraftingAI Attack Adaptation · AML.T0066
- 2LLM Prompt CraftingAI Attack Adaptation · AML.T0065
- 3Valid AccountsInitial Access · AML.T0012
- 4AI-Enabled Product or ServiceAI Model Access · AML.T0047
- 5RAG PoisoningPersistence · AML.T0070
- 6LLM Prompt Injection: IndirectExecution · AML.T0051.001
- 7RAG Credential HarvestingCredential Access · AML.T0082
- 8LLM Response RenderingExfiltration · AML.T0077
How these attacks work
Software used
- Apktool · Tool
From MITRE ATLAS 2026.10 (Apache-2.0). ShieldMap lists who and what was involved and the techniques used, and links to ATLAS for the write-up; it does not copy the steps. Attribution is ATLAS's, not ShieldPi's.