---
format: "aidr-story-markdown/v1"
id: "9d5e5475de700473f713e091a376ab58170185e5b6c2f40a3017a9e40d6996ed"
canonical_url: "https://aidr.today/9d5e5475?lang=en"
title: "A single prompt was enough to hijack every AI agent in an AWS account, Zenity researchers found"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-08T13:01:01.000Z"
category: "Agents"
topics: ["amazon","agent"]
source_urls: ["https://the-decoder.com/a-single-prompt-was-enough-to-hijack-every-ai-agent-in-an-aws-account-zenity-researchers-found/"]
summary: "Zenity Labs researchers found that a single publicly accessible AI agent on Amazon's Bedrock AgentCore could be used to take over every AgentCore agent in the same AWS account and region. The attack exploited an internal AWS interface for temporary cloud credentials, but AWS has since patched the issue and tightened default agent permissions."
---

# A single prompt was enough to hijack every AI agent in an AWS account, Zenity researchers found

> [Open the canonical story](<https://aidr.today/9d5e5475?lang=en>)

**Published:** 2026-10-08T13:01:01.000Z
**Category:** Agents
**Topics:** amazon, agent

## Summary

Zenity Labs researchers found that a single publicly accessible AI agent on Amazon's Bedrock AgentCore could be used to take over every AgentCore agent in the same AWS account and region\. The attack exploited an internal AWS interface for temporary cloud credentials, but AWS has since patched the issue and tightened default agent permissions\.

## Sources

- [Story source](<https://the-decoder.com/a-single-prompt-was-enough-to-hijack-every-ai-agent-in-an-aws-account-zenity-researchers-found/>)