--- title: Roblox downloaded 145GB models from us (including adult... description: A crawler identifying itself as Roblox's downloaded 145.5 GB of 3D models from Open3DLab, SFMLab and SmutBase without permission. Over 15 and 16 October 2026... canonical: https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Roblox downloaded 145GB models from us (including adult content) without asking | daily.dev og:description: A crawler identifying itself as Roblox's downloaded 145.5 GB of 3D models from Open3DLab, SFMLab and SmutBase without permission. Over 15 and 16 October 2026... og:url: https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x og:image: https://api.daily.dev/og/posts/hpmZ8TS2x.png og:image:alt: Roblox downloaded 145GB models from us (including adult content) without asking og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Roblox downloaded 145GB models from us (including adult content) without asking **[Hacker News](https://daily.dev/sources/hn)** · 14 min read · 0 upvotes · 0 comments ## Summary A crawler identifying itself as Roblox's downloaded 145.5 GB of 3D models from Open3DLab, SFMLab and SmutBase without permission. Over 15 and 16 October 2026 it pulled 1,027 files from 712 projects after reading nearly the entire catalogue of 18,926 projects. The user agent read 'trellis2-data-sourcing/1.0' and listed a Roblox researcher's email tied to Roblox's Cube 3D foundation model work. It ignored robots.txt, the sites' Terms of Service, and NonCommercial licenses on 423 of the projects. 345 of the 712 downloaded projects were adult content, taken at the same rate as everything else, with no apparent filtering. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How much data did Roblox's AI crawler download from 3D model hosting sites? A crawler using the user agent 'trellis2-data-sourcing/1.0' downloaded 145.5 GB across 1,027 files from 712 projects on Open3DLab, SFMLab and SmutBase over 15 and 16 October 2026. It had first read the main pages of 18,926 projects, nearly the entire catalogue, in 26,429 requests over 34 hours from a single AWS address. _Developers tracking AI scraping incidents and data provenance disputes can follow cases like this on daily.dev._ ### Did the Roblox scraping crawler download adult content and ignore licenses? Yes, 345 of the 712 scraped projects were adult content, taken at the same rate as non-adult projects, including 224 projects from SmutBase, a site exclusively for adult material. The crawler also ignored the sites' robots.txt and Terms of Service, and defied licenses on 423 of the 712 projects that were marked for non-commercial use only. _Anyone weighing how to license creative work against AI scrapers can track developments like this on daily.dev._ ### What is Roblox's Cube 3D foundation model and how does it relate to this data scraping incident? Cube is Roblox's generative AI system for 3D and 4D content, introduced in March 2025, intended to eventually handle character rigging and avatar generation. The email address embedded in the scraping crawler's user agent belonged to a researcher listed among the 40-plus authors of the Cube research paper, suggesting a link between the crawler and Roblox's 3D foundation model work, though Roblox did not confirm this. _Engineers evaluating generative 3D AI tools and their training data sourcing can follow this story on daily.dev._ ## Community take How the wider developer community reacted, aggregated from 2 discussions and 165 comments across hackernews (as of 2026-10-28). **TL;DR:** Discussion largely sides with the hosting sites against Roblox's crawler, though a significant contingent argues the hosts themselves profit from infringing fan-made content, and there's heated legal debate over whether downloading/training actually violates copyright or licenses. **Sentiment:** 5% positive · 25% mixed · 70% skeptical **The case for** - Some argue downloading publicly accessible files isn't inherently a copyright or legal violation. - A few note training on scraped data may currently be legal regardless of license terms. - Some see this as unsurprising/standard practice given how data scraping already works on the web. **The pushback** - Many call out the hypocrisy of sites hosting ripped/infringing game assets complaining about being scraped themselves. - Commenters argue ignoring robots.txt, ToS, and explicit non-commercial licenses is still wrong even if technically hard to enforce. - Some see this as part of a broader pattern of large companies treating scraping as a cost of doing business while smaller operators bear the server/infrastructure burden. - Frustration that megacorporations can enforce their own ToS legally while being accused of ignoring others'. **By community** - hackernews (heated): Debate split between condemning Roblox's disregard for licenses/robots.txt and pointing out the hosting sites' own questionable hosting of ripped game assets, with a long tangent on blocking datacenter IPs to fight scrapers. **Hottest debate:** Whether downloading/training on licensed content actually constitutes a legal or moral violation, versus the counter-argument that the hosting sites themselves profit from copyright-infringing fan content. **Open questions** - Will courts ultimately decide that AI training on scraped/licensed content counts as fair use or derivative infringement? - Can non-commercial license terms actually be enforced against a company that merely downloaded (rather than redistributed) the files? **Highlights** > Party A: hosts largely copyrighted material, says they are opposed to copyright in its current form Party B: massive IP business , very aggressive about everything from brand to software, defends and monetizes current copyright regime I think it’s fair for party A to call out party B for downloading a massive trove of known-copyrighted material. > — [brookst on hackernews](https://news.ycombinator.com/item?id=50257964) > If you read the article, you'd find they admit to that, their position is nuanced, and this is not a "gotcha". > We'll be honest about where we stand. We don't like copyright as it exists. We'd like to see it radically reformed, and we wouldn't mourn it if it were abolished. Our community lives in its grey areas every day: many of the models on our sites are fan ports of characters from commercial games, made out of love and shared for free. When a studio sends us a takedown notice, we comply. > While we as small artists and fan communities get copyright used against us, the world's largest companies are committing copyright infringement at a scale unseen in human history. This has been the case since the invention of the internet of course; digital media make duplication, remixing and resharing trivial, but copyright law never caught up. Filesharing stabbed copyright in the gut. AI is just twisting the knife, and completes the creation an unequal system where copyright infringement is only a crime if you're a low net worth individual. Rich companies are infringing copyright at industrial scale in order to build their LLMs, and even if caught and challenged, they don't suffer any catastrophic punishment. The infringed party gets enough money to shut them up and everyone moves on. A fine is a price. They'll keep doing it with impunity. Meanwhile the small guy who doesn't have copyright lawyers on retainer is getting shafted, time and time again. > — [amiga386 on hackernews · 2 comments](https://news.ycombinator.com/item?id=50257852) > It's a community site, first model I looked at (Tanya by Shrimport) is under CC-BY-ND-NC licence (Creative Commons, By attribution, No-derivatives, Non-commercial). AI training by definition is derivative, and Roblox not using it for commercial purposes if used in production is practically nil. This is a huge part of why AI upsets people, we can't have nice things such as communities where people are sharing 3d models without it being feeding ground for commercial companies that has little intention to ever honor any licences or compensate authors. > — [whizzter on hackernews · 1 comments](https://news.ycombinator.com/item?id=50257608) > How can you start an outcry like this, but host stuff like this, tagged with a permissive license: https://open3dlab.com/project/e2a0830d-f166-4b1b-aabf-8ca735... Just an example, there's plenty more on the frontpage. I have nothing against these guys, but I just can't understand the moral high ground of the complaint. > — [phn on hackernews · 8 comments](https://news.ycombinator.com/item?id=50257622) > Tons of those are (as the site admins admit) models that are themselves infringing on copyright, and thus any license based on copyright ownership is invalid. You can't enforce a license on a 3d model you in turn stole from a game. > — [ang\_cire on hackernews](https://news.ycombinator.com/item?id=50258603) **Source threads** - [hackernews](https://news.ycombinator.com/item?id=50256985) · 235 points · 165 comments - [hackernews](https://news.ycombinator.com/item?id=50253348) · 8 points · 0 comments ## Similar posts on daily.dev - [State of Open Source on Hugging Face: Spring 2026](https://daily.dev/posts/state-of-open-source-on-hugging-face-spring-2026-fuzipfbpa) · Hugging Face · 2 upvotes · 0 comments - [Malicious Hugging Face model masquerading as OpenAI release hits 244K downloads](https://daily.dev/posts/malicious-hugging-face-model-masquerading-as-openai-release-hits-244k-downloads-xftpmcpz1) · CSO Online · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#genai](https://daily.dev/tags/genai), [#crawling](https://daily.dev/tags/crawling) [View this post on daily.dev](https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Roblox downloaded 145GB models from us (including adult content) without asking","url":"https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x"},"datePublished":"2026-10-27T11:50:41.486Z","dateModified":"2026-10-28T15:58:46.427Z","description":"A crawler identifying itself as Roblox's downloaded 145.5 GB of 3D models from Open3DLab, SFMLab and SmutBase without permission. Over 15 and 16 October 2026...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/69de4cb355d7f34ee29e6a6c0e0a36f3?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/69de4cb355d7f34ee29e6a6c0e0a36f3?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,genai,crawling","timeRequired":"PT14M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Roblox downloaded 145GB models from us (including adult content) without asking"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/roblox-downloaded-145gb-models-from-us-including-adult-content-without-asking-hpmz8ts2x#faq","mainEntity":[{"@type":"Question","name":"How much data did Roblox's AI crawler download from 3D model hosting sites?","acceptedAnswer":{"@type":"Answer","text":"A crawler using the user agent 'trellis2-data-sourcing/1.0' downloaded 145.5 GB across 1,027 files from 712 projects on Open3DLab, SFMLab and SmutBase over 15 and 16 October 2026. It had first read the main pages of 18,926 projects, nearly the entire catalogue, in 26,429 requests over 34 hours from a single AWS address. Developers tracking AI scraping incidents and data provenance disputes can follow cases like this on daily.dev."}},{"@type":"Question","name":"Did the Roblox scraping crawler download adult content and ignore licenses?","acceptedAnswer":{"@type":"Answer","text":"Yes, 345 of the 712 scraped projects were adult content, taken at the same rate as non-adult projects, including 224 projects from SmutBase, a site exclusively for adult material. The crawler also ignored the sites' robots.txt and Terms of Service, and defied licenses on 423 of the 712 projects that were marked for non-commercial use only. Anyone weighing how to license creative work against AI scrapers can track developments like this on daily.dev."}},{"@type":"Question","name":"What is Roblox's Cube 3D foundation model and how does it relate to this data scraping incident?","acceptedAnswer":{"@type":"Answer","text":"Cube is Roblox's generative AI system for 3D and 4D content, introduced in March 2025, intended to eventually handle character rigging and avatar generation. The email address embedded in the scraping crawler's user agent belonged to a researcher listed among the 40-plus authors of the Cube research paper, suggesting a link between the crawler and Roblox's 3D foundation model work, though Roblox did not confirm this. Engineers evaluating generative 3D AI tools and their training data sourcing can follow this story on daily.dev."}}]} ```