GCD AI RADAR
incidents/AIR-2025-0061

Data Exfiltration from Slack AI via Indirect Prompt Injection

Attack on AI1 Research / proof of conceptPrompt injectionAgent hijackData leak

Researchers demonstrated that private data can be exfiltrated from Slack AI using indirect prompt injection via public channel posts. The attack caused the AI to retrieve and execute malicious instructions, leaking sensitive API keys.

Details

type
Attack on AI
An AI model, agent or AI-powered product was the thing exploited, abused or leaked.
categories
  • Prompt injectionprimary
    Untrusted content (web pages, documents, emails, tool output) steers a model into ignoring its instructions.
  • Agent hijack
    An AI agent is tricked into unauthorized actions: tool misuse, data exfiltration through the agent, or acting against its principal.
  • Data leak
    Training data, chat logs, system prompts, model weights or other AI-system data exposed.
owasp llm top 10
mitre atlas
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
1 · Research / proof of concept
Demonstrated by researchers; no known real-world exploitation.
targets
Enterprise deployment
products
Slack AI
incident date
20 Aug 2024
first reported
7 Nov 2025
origin
MITRE ATLAS
Derived from MITRE ATLAS (Apache 2.0, © MITRE).
classifier confidence
0.95 · published at or above 0.80; see the methodology
tags
prompt-injection · data-exfiltration · rag · slack-ai
record
created 2026-10-18 06:55Z · updated 2026-10-18 06:55Z · json

Sources · 2

  1. 2025-11-07
    Data Exfiltration from Slack AI via Indirect Prompt Injection
    MITRE ATLAS · atlas.mitre.org/studies/AML.CS0035
  2. 2025-11-07
    Data Exfiltration from Slack AI via indirect prompt injection
    promptarmor.substack.com · promptarmor.substack.com/p/data-exfiltration-from-slack-ai-via

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents