---
format: "aidr-story-markdown/v1"
id: "b056934f9faee5195eb828e9d4bbf7c9e851b3ae4d2415f035b11ef8ba9d0313"
canonical_url: "https://aidr.today/b056934f?lang=en"
title: "An AI agent in \"Skynet mode\" hijacks unprotected language models to launch self-sustaining cyberattacks"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-22T11:30:56.000Z"
category: "Agents"
topics: ["agent"]
source_urls: ["https://the-decoder.com/an-ai-agent-in-skynet-mode-hijacks-unprotected-language-models-to-launch-self-sustaining-cyberattacks/"]
summary: "Security researchers at Symantec have found an offensive AI agent that hijacks unprotected language models on the web to attack servers, steal databases, and install cryptominers without human input, running up to 100,000 steps autonomously. The owner of the hijacked model reportedly gets stuck with the bill."
---

# An AI agent in "Skynet mode" hijacks unprotected language models to launch self\-sustaining cyberattacks

> [Open the canonical story](<https://aidr.today/b056934f?lang=en>)

**Published:** 2026-10-22T11:30:56.000Z
**Category:** Agents
**Topics:** agent

## Summary

Security researchers at Symantec have found an offensive AI agent that hijacks unprotected language models on the web to attack servers, steal databases, and install cryptominers without human input, running up to 100,000 steps autonomously\. The owner of the hijacked model reportedly gets stuck with the bill\.

## Sources

- [Story source](<https://the-decoder.com/an-ai-agent-in-skynet-mode-hijacks-unprotected-language-models-to-launch-self-sustaining-cyberattacks/>)