PulseAugur
EN
LIVE 10:51:58

AI agents tricked into running malware via website instructions · 1 source tracked

By PulseAugur Editorial · [1 sources] ·

Researchers have discovered a significant vulnerability in how AI agents interact with company websites, specifically through the use of `llms.txt` files. These files, intended to guide AI agents on website interaction, can be manipulated to execute arbitrary code. The researchers successfully tricked AI agents from Fortune 500 companies into running malicious code by exploiting outdated or non-existent package references within these `llms.txt` files. This highlights a new frontier in supply-chain attacks where data, in the form of instructions, can directly lead to code execution. AI

IMPACT This vulnerability highlights a new attack vector for AI agents, potentially impacting enterprise security and the integrity of software supply chains.

RANK_REASON The article details a security vulnerability and exploit method affecting AI agents and software supply chains, but does not announce a new model release or core research breakthrough from a frontier lab.

Read on Tom's Hardware →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

AI agents tricked into running malware via website instructions · 1 source tracked

How we ranked this

Signal score
62 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The article details a security vulnerability and exploit method affecting AI agents and software supply chains, but does not announce a new model release or core research breakthrough from a fronti…
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
product, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. Tom's Hardware TIER_1 English(EN) · Bruno Ferreira ·

    Researchers easily trick Fortune-500 companies' AI agents into running arbitrary code — supply-chain attack via llms.txt guidance file illustrates how data has become code

    Researchers easily trick Fortune-500 companies' AI agents into running arbitrary code. This supply-chain attack, done via using data in public llms.txt guidance files, illustrates the dangers of data becoming code.

RELATED ENTITIES

RELATED TOPICS