GCD AI RADAR
incidents/AIR-2026-0340

Data Destruction via Indirect Prompt Injection in Claude Computer-Use

Attack on AI1 Research / proof of conceptPrompt injectionAgent hijack

Researchers demonstrated that an indirect prompt injection embedded in a PDF file could trick Claude Computer-Use into invoking bash commands. The attack resulted in simulated deletion of the victim filesystem.

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.
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
Anthropic, Framework / tooling
products
Claude Computer-Use
incident date
24 Oct 2024
first reported
30 Jan 2026
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 · agent_hijack · computer_use · hiddenlayer
record
created 2026-10-18 06:55Z · updated 2026-10-18 06:55Z · json

Sources · 2

  1. 2026-01-30
    Data Destruction via Indirect Prompt Injection Targeting Claude Computer-Use
    MITRE ATLAS · atlas.mitre.org/studies/AML.CS0046
  2. 2026-01-30
    Indirect Prompt Injection of Claude Computer Use
    HiddenLayer · hiddenlayer.com/innovation-hub/indirect-prompt-injection-of-claude-computer-use

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