Meta Sued Over Training Data for Its AI and Face-Recognition Systems
Meta Sued Over Training Data for Its AI and Face-Recognition Systems
Meta 因其 AI 及人脸识别系统的训练数据遭起诉
A set of parents and their children in Illinois and California filed a lawsuit last week in federal court in Chicago alleging that Meta illegally used their Facebook and Instagram photos to build NameTag, an unreleased face-recognition system for its smart glasses, and to train generative AI models including Emu and Muse Image.
上周，来自伊利诺伊州和加利福尼亚州的一群家长及其子女在芝加哥联邦法院提起诉讼，指控 Meta 非法使用他们的 Facebook 和 Instagram 照片来构建 NameTag（一种尚未发布的智能眼镜人脸识别系统），并用于训练包括 Emu 和 Muse Image 在内的生成式 AI 模型。
The proposed class action alleges that Meta violated Illinois and California privacy laws by extracting biometric information from people’s photos without notice or consent.
这项拟议的集体诉讼指控 Meta 在未经通知或同意的情况下，从人们的照片中提取生物识别信息，违反了伊利诺伊州和加利福尼亚州的隐私法。
WIRED reported in June that code for NameTag had been secretly embedded in the Meta glasses AI companion app, which had been downloaded more than 50 million times. While the feature had not been enabled for users of the app, the analysis found that the system was designed to turn faces captured by the glasses into biometric signatures and compare them with so-called faceprints stored in a database on the user’s phone. That database was configured to receive updates from Meta. At the time, WIRED could not determine where the underlying faceprint data came from.
《连线》（WIRED）在六月份报道称，NameTag 的代码被秘密嵌入到 Meta 眼镜的 AI 伴侣应用中，该应用下载量已超过 5000 万次。虽然该功能并未对用户启用，但分析发现，该系统的设计初衷是将眼镜捕捉到的人脸转化为生物识别特征，并将其与存储在用户手机数据库中的所谓“人脸指纹”（faceprints）进行比对。该数据库被配置为可接收来自 Meta 的更新。当时，《连线》无法确定这些底层人脸指纹数据的来源。
The complaint alleges that those faceprints are possibly derived from Facebook and Instagram images, citing reporting that Meta employees claimed that NameTag could recognize people through their Meta connections or public Instagram accounts, along with a company patent describing face matching against profile photos and other images held by Meta.
诉状称，这些人脸指纹可能源自 Facebook 和 Instagram 的图片。诉状援引相关报道称，Meta 员工曾表示 NameTag 可以通过 Meta 社交关系或公开的 Instagram 账户识别用户，此外公司的一项专利也描述了将人脸与 Meta 持有的个人资料照片及其他图像进行比对的技术。
Meta told WIRED in June that it was “not building a central face database,” but it would not answer questions about whether NameTag would be opt-in or how the system would retain faceprints. The complaint acknowledges that Meta has not disclosed which images, if any, were used for generating biometric data, saying that information remains solely in the company’s possession.
Meta 在六月份告诉《连线》，公司“并未建立中央人脸数据库”，但拒绝回答关于 NameTag 是否会采用“选择加入”（opt-in）机制，以及系统将如何保留人脸指纹等问题。诉状承认，Meta 尚未披露哪些图像（如果有的话）被用于生成生物识别数据，并指出这些信息仍完全掌握在公司手中。
The suit also targets Meta’s image-generation systems. Meta has said it trained Emu on large quantities of Facebook and Instagram images and text, with chief product officer Chris Cox calling those platforms a “data advantage” for its AI systems. The complaint alleges that the training process illegally harvested biometric information about people who appeared in the images. Muse Image, released this summer, had earlier drawn criticism after allowing users to generate images based on other people’s public Instagram accounts, a feature the company removed within days after saying it had “missed the mark.”
该诉讼还针对 Meta 的图像生成系统。Meta 曾表示，其 Emu 模型是在海量的 Facebook 和 Instagram 图片及文本上训练出来的，首席产品官 Chris Cox 将这些平台称为其 AI 系统的“数据优势”。诉状指控该训练过程非法收集了出现在图片中的人的生物识别信息。今年夏天发布的 Muse Image 此前曾引发争议，因为它允许用户根据他人的公开 Instagram 账户生成图像；在公司表示该功能“未达预期”后，于几天内将其移除。
“This lawsuit is without merit and misrepresents our work. We’ve been transparent about how we use people’s information to build and improve our AI products. As for NameTags, nothing has shipped to consumers and no final decision has been made on what to do here, if anything,” a Meta spokesperson says in a statement. “If we do decide to roll something out, we will take a thoughtful approach and do so with full transparency. One decision we can be clear about—we are not building a universal face database.”
Meta 发言人在声明中表示：“这起诉讼毫无根据，歪曲了我们的工作。我们对于如何使用用户信息来构建和改进 AI 产品一直保持透明。至于 NameTag，目前没有任何产品交付给消费者，对于是否推进该项目也尚未做出最终决定。如果我们决定推出相关功能，我们将采取审慎的态度，并确保完全透明。有一点我们可以明确：我们并没有在建立通用的面部数据库。”
“People shouldn’t have to worry if their biometric information will be misused simply because their photographs appear on a social media platform,” Justin Boley, a partner at Wexler Boley & Elgersma and an attorney for the plaintiffs, said in a statement.
原告律师、Wexler Boley & Elgersma 合伙人 Justin Boley 在声明中表示：“人们不应该仅仅因为照片出现在社交媒体平台上，就担心自己的生物识别信息会被滥用。”
The plaintiffs are Francisco Alvarez and his son, both Illinois residents, and Jeremy Wahl, a California resident, and his 10-year-old daughter. But the proposed class includes people in Illinois, California, and across the United States whose images were uploaded to Facebook or Instagram or were submitted to Meta’s generative AI systems through prompts, dating back to September 4, 2021. The complaint estimates that the national class could number in the millions.
原告包括伊利诺伊州居民 Francisco Alvarez 及其儿子，以及加利福尼亚州居民 Jeremy Wahl 及其 10 岁的女儿。但拟议的集体诉讼涵盖了自 2021 年 9 月 4 日起，所有在伊利诺伊州、加利福尼亚州及全美范围内，其图像被上传至 Facebook 或 Instagram，或通过提示词提交给 Meta 生成式 AI 系统的用户。诉状估计，全国范围内的集体诉讼人数可能达到数百万。
Under Illinois’ Biometric Information Privacy Act, the plaintiffs are seeking $5,000 for each intentional or reckless violation, or actual damages if greater, and $1,000 for each negligent violation, or actual damages if greater, as well as injunctive relief. The California claims seek additional damages and other relief.
根据伊利诺伊州的《生物识别信息隐私法》，原告要求对每一次故意或鲁莽的违规行为赔偿 5,000 美元（或实际损失，以较高者为准），对每一次过失违规行为赔偿 1,000 美元（或实际损失，以较高者为准），并寻求禁令救济。加利福尼亚州的诉讼请求则寻求额外的赔偿及其他救济。
This is not the first time Meta has faced fines over its handling of biometric data. In 2020, the company agreed to pay $650 million to settle an Illinois class action over an earlier face-recognition system. In November 2021, the company announced that it would shut down the system and delete more than a billion faceprints. In 2024, Meta agreed to pay Texas $1.4 billion to resolve separate allegations that it had unlawfully collected biometric data from users.
这并非 Meta 首次因处理生物识别数据而面临罚款。2020 年，该公司同意支付 6.5 亿美元，以和解一起关于早期人脸识别系统的伊利诺伊州集体诉讼。2021 年 11 月，该公司宣布将关闭该系统并删除超过 10 亿条人脸指纹。2024 年，Meta 同意向德克萨斯州支付 14 亿美元，以解决关于其非法收集用户生物识别数据的另一起指控。
The day after WIRED’s June 4 report about NameTag, Meta removed the code from its app. The company argued that the feature never existed because it was not available to consumers, even though WIRED’s analysis and testing by outside researchers found a technically functional face-recognition system shipped inside an app downloaded by tens of millions of people.
在《连线》6 月 4 日关于 NameTag 的报道发布后的第二天，Meta 从其应用中移除了相关代码。该公司辩称该功能从未存在，因为它并未向消费者开放，尽管《连线》的分析和外部研究人员的测试发现，一个技术上可运行的人脸识别系统确实被内置在了一个被数千万人下载的应用中。
Meta CTO Andrew Bosworth called WIRED’s reporting “incredibly misleading” and “absolutely dishonest.” Weeks later, Bosworth described NameTag on a podcast, saying it could recognize people that a glasses wearer had previously met and had asked the device to remember. “I think [it] would be a great feature,” he said. Meta continued to describe NameTag as something it was exploring rather than a product available to consumers.
Meta 首席技术官 Andrew Bosworth 称《连线》的报道“极具误导性”且“极其不诚实”。几周后，Bosworth 在播客中描述了 NameTag，称它可以识别眼镜佩戴者之前见过并要求设备记住的人。他说：“我认为这会是一个很棒的功能。”Meta 继续将 NameTag 描述为正在探索的项目，而非向消费者提供的产品。
The complaint frames the case as part of a much longer pattern of privacy violations from Meta. Reaching back into Facebook’s earliest days, it cites a 2004 chat in which CEO Mark Zuckerberg reportedly referred to people who had trusted him with their data as “dumb fucks.”
诉状将此案定性为 Meta 长期以来一系列隐私违规行为的一部分。诉状追溯到 Facebook 成立初期，引用了 2004 年的一段聊天记录，据报道，首席执行官马克·扎克伯格在其中将那些信任他并提供数据的用户称为“蠢货”（dumb fucks）。
Updated 9/11/2026 at 3:09 pm EDT: Added comment from Meta. 更新于美国东部时间 2026 年 9 月 11 日下午 3:09：添加了来自 Meta 的评论。