Roblox Crawler Accused of Scraping Community 3D Models

A data crawler allegedly identified as Roblox reportedly scraped 145.5 GB of 3D model files without permission.

Updated on Oct. 27, 2026 in Artificial Intelligence

Isometric editorial illustration of abstract 3D geometric prisms floating in a structured space, representing unauthorized data extraction. — Operators of three community 3D model websites have blocked a Roblox-linked crawler that allegedly scraped over 145 GB of asset data. AI Illustration. Upload story photo >

Live Poll

Should companies be prohibited from using non-commercial content to train AI models without explicit permission?

Three community websites report that a crawler identifying itself as Roblox downloaded over 1,000 files from their platforms. Operators suspect the activity was intended to gather training data for an AI model.

Why it matters

The incident highlights growing concerns over how companies source data for AI development, particularly when scrapers bypass site terms of service and ignore robots.txt instructions. Many of the accessed projects were explicitly licensed for non-commercial use only.

The crawler performed 26,429 requests across 18,926 project pages over a 34-hour window. This activity resulted in the unauthorized download of 1,027 files across 712 distinct projects, including 345 files containing adult content.

The players

Roblox

An online game platform and game creation system that allows users to program games and play games created by other users.

Open3DLab

A community-driven website that hosts 3D models and digital assets created by independent developers.

SFMLab

An online platform serving as a repository for 3D model sharing and digital art collaboration.

SmutBase

An online community platform that provides hosting services for adult-oriented 3D content and digital assets.

AWS

A comprehensive cloud computing platform provided by Amazon that offers various infrastructure services to businesses and developers.

The details

The crawler utilized a user agent linked to a Roblox email address and originated from an AWS cloud environment. Website operators at Open3DLab, SFMLab, and SmutBase blocked the access attempts after identifying the unauthorized collection of their hosted assets.

October 15-16, 2026: Crawler downloaded 1,027 files from three community websites.
October 18, 2026: Site operators discovered the activity and initiated a block.
October 19, 2026: Crawler returned and was blocked by site operators.
October 20, 2026: Crawler activity ceased at approximately 15:00.
October 26, 2026: Operator published a detailed account of the incident.

The Tech Race

The use of automated crawlers to harvest massive datasets represents the aggressive data-acquisition strategies currently defining the AI arms race. This move highlights a growing friction between web-based content creators and the massive ingestion requirements of modern machine learning models.

Users of 3D asset platforms may face increased restrictions on how their creations are shared as sites implement more aggressive blocking mechanisms to protect data. Developers using these sites must now be more vigilant about licensing terms to ensure their work remains protected from bulk collection.

The takeaway

This case illustrates the ongoing tension between independent content creators and automated data harvesting in the AI era. Site operators should regularly monitor server logs to identify and mitigate unauthorized large-scale data ingestion.