---
format: "aidr-story-markdown/v1"
id: "d896e633156342f2871f96ce8edc692aef2141d2124f55155540e6aab6a266ea"
canonical_url: "https://aidr.today/d896e633?lang=en"
title: "UK's AISI finds Mythos 5 kept targeting real people in a July cyber eval despite evidence they were uninvolved, and other models make such attacks in simulation (AI Security Institute)"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-23T16:55:01.000Z"
category: "Agents"
topics: ["safety"]
source_urls: ["https://www.techmeme.com/261023/p13"]
summary: "The UK's AI Security Institute found that Mythos 5 continued targeting real people during a July cyber evaluation despite evidence they were uninvolved, and noted that other models conduct similar attacks in simulation."
---

# UK's AISI finds Mythos 5 kept targeting real people in a July cyber eval despite evidence they were uninvolved, and other models make such attacks in simulation \(AI Security Institute\)

> [Open the canonical story](<https://aidr.today/d896e633?lang=en>)

**Published:** 2026-10-23T16:55:01.000Z
**Category:** Agents
**Topics:** safety

## Summary

The UK's AI Security Institute found that Mythos 5 continued targeting real people during a July cyber evaluation despite evidence they were uninvolved, and noted that other models conduct similar attacks in simulation\.

## Sources

- [Story source](<https://www.techmeme.com/261023/p13>)