---
title: "Hidden PDF text can hijack Atlassian's AI agent Rovo"
url: https://www.parallelquant.com/posts/hidden-pdf-text-can-hijack-atlassian-s-ai-agent-rovo-7acfaf
source_name: "The Decoder"
source_url: https://the-decoder.com/hidden-text-in-a-pdf-is-enough-to-steal-sensitive-data-through-atlassians-ai-agent-rovo/
published: 2026-08-10T08:46:36.000Z
topics: ["security", "agents"]
publisher: "Parallel Quant"
---

# Hidden PDF text can hijack Atlassian's AI agent Rovo

*2026-08-10 · Source: [The Decoder](https://the-decoder.com/hidden-text-in-a-pdf-is-enough-to-steal-sensitive-data-through-atlassians-ai-agent-rovo/)*

Security firm PromptArmor showed that hidden instructions embedded in a PDF can hijack Atlassian's Rovo AI agent, silently forwarding sensitive data from Jira and Confluence to an external server. The attack requires no user confirmation and leaves no visible trace.

**Why it matters:** This is another concrete instance of indirect prompt injection, where the malicious instructions arrive through a document rather than a chat message, a class of attack labs have struggled to fully close off. It adds to a growing string of real-world agent security incidents as tools like Rovo get deeper access to internal company data.

**Topics:** security, agents

---
Read the original: https://the-decoder.com/hidden-text-in-a-pdf-is-enough-to-steal-sensitive-data-through-atlassians-ai-agent-rovo/
Canonical: https://www.parallelquant.com/posts/hidden-pdf-text-can-hijack-atlassian-s-ai-agent-rovo-7acfaf
Published by Parallel Quant — https://www.parallelquant.com