游戏公司Roblox于2026年10月15日至16日期间，通过名为“trellis2-data-sourcing”的用户代理，从Open3DLab、SFMLab及SmutBase等网站未经授权使用爬虫批量下载了总计145.5 GB的3D资产文件 1。此次抓取行为涉及712个项目的1,027个具体文件，主要发生于上述两天，并于10月18日被受害方发现并阻断 1。值得注意的是，所获取的数据中包含大量敏感内容，具体包括来自SmutBase的364个成人内容文件（共54.8 GB），以及另外345个被标记为成人内容的文件（共62.6 GB） 1。
Roblox此举因严重违反网络规范而引发社区强烈批评，其行为无视了目标网站的robots.txt协议和服务条款 1。此外，在涉及的423个项目中，Roblox还违反了创作者设定的“非商业使用”许可证规定 1。尽管事件曝光后受害网站向Roblox发送邮件问询，但截至目前，该公司未对相关指控作出任何回应 1。外界推测，这些被非法抓取的3D模型可能被用于训练Roblox旗下的Cube 3D基础模型或与Microsoft TRELLIS.2相关的AI项目 1。这一事件再次加剧了关于大型科技公司滥用数据进行AI训练的伦理与法律争论，此前Meta、Nvidia和Anthropic等公司也曾面临类似的未经授权大规模抓取数据的指控 1。
Gaming company Roblox faces community criticism after being accused of unauthorized data scraping from multiple online repositories, including Open3DLab, SFMLab, and SmutBase. Reports indicate that between October 15 and 16, 2026, the company used a web crawler identified by the user agent "trellis2-data-sourcing" to download approximately 145.5 GB of 3D assets 1. The activity was discovered and subsequently blocked on October 18, 2026 1.
The alleged infringement involved 1,027 files originating from 712 distinct projects 1. A significant portion of this data included sensitive material: specifically, 364 adult content files totaling 54.8 GB were sourced directly from SmutBase, while an additional 345 projects containing 62.6 GB of data were flagged as adult content 1. Critics argue that these downloads violate the terms of service of the hosting platforms and ignore established robots.txt protocols 1. Furthermore, the extraction reportedly disregarded non-commercial use licenses attached to 423 of the downloaded projects 1.
Roblox has not responded to email inquiries from the affected websites following the exposure of these activities 1. The incident has intensified ongoing debates regarding the ethical and legal implications of large technology companies utilizing scraped data for AI training 1. The 3D assets are suspected to have been intended for training Roblox’s Cube 3D foundation model or related Microsoft TRELLIS.2 projects 1. This allegation follows similar accusations against other major tech firms, such as Meta, Nvidia, and Anthropic, which have previously faced scrutiny for unauthorized large-scale data collection for artificial intelligence development 1.
评论
还没有评论，欢迎留下第一条。