---
id: 6760
title: "Inside 90 days of attacks on AI infrastructure"
source_domain: wiz.io
source_url: "https://www.wiz.io/blog/ai-infrastructure-honeypot"
published_at: "2026-08-27T16:33:16Z"
tags: ["ai-security", "cloud", "rce", "threat-intelligence", "mcp", "prompt-injection", "llm-security", "security-research", "threat-research", "litellm", "honeypot", "wiz"]
phying_url: "https://phying.com/item/6760"
lang: en
---

# Inside 90 days of attacks on AI infrastructure

## Summary

Wiz honeypots logged 90 days of attacks on AI infrastructure — LiteLLM, Flowise, LangChain, ChromaDB, Ollama — via MCP server RCE, blind prompt injection, and AI-native post-exploitation tooling.

## Why it matters

This research provides real-world threat telemetry on how attackers target modern AI deployments and MCP interfaces.

## Intelligence Metadata

- **Sources:** wiz.io
- **Published:** 2026-08-27T16:33:16Z
- **Original:** https://www.wiz.io/blog/ai-infrastructure-honeypot
- **Phying News:** https://phying.com/item/6760