Meta Silently Added Face-Recognition Code for Its Smart Glasses to Millions of Phones
Meta Silently Added Face-Recognition Code for Its Smart Glasses to Millions of Phones
Meta 在数百万部手机中静默植入智能眼镜人脸识别代码
Meta has quietly embedded face-recognition technology for its smart glasses into an app downloaded to millions of phones, according to a WIRED analysis of the company’s software. 根据《连线》（WIRED）对该公司软件的分析，Meta 已将其智能眼镜的人脸识别技术静默植入到一款下载量达数百万次的应用程序中。
Code discreetly added to Meta’s AI app over multiple updates this year shows that the feature, internally called “NameTag,” identifies people captured by the glasses’ camera and, when activated, alerts the wearer when it recognizes someone. 今年通过多次更新悄悄添加到 Meta AI 应用中的代码显示，该功能在内部被称为“NameTag”。它能够识别眼镜摄像头捕捉到的人，并在激活后，当识别出某人时向佩戴者发出提醒。
The discovery of NameTag in the live Meta AI app shows that Meta had begun shipping face-recognition code to users’ phones while publicly describing it as something the company was still “thinking through.” In April, Meta said if it were to utilize face recognition, it wouldn’t be rolled out without first taking “a very thoughtful approach.” But WIRED found that as early as January, core components of the system had been integrated into software distributed to millions of people. 在现行的 Meta AI 应用中发现 NameTag 表明，Meta 在公开宣称该技术仍处于“思考阶段”的同时，就已经开始向用户的手机推送人脸识别代码。今年 4 月，Meta 曾表示，如果使用人脸识别，在未采取“非常审慎的方法”之前不会推出。但《连线》发现，早在 1 月份，该系统的核心组件就已经集成到分发给数百万人的软件中了。
Though not yet enabled, NameTag sits inside a Meta AI companion app that’s been downloaded over 50 million times and is necessary for use of key features of its smart glasses, including Ray-Ban and Oakley models. If activated, it will transform faces captured by Meta’s glasses into unique biometric signatures, commonly known as faceprints, and check each one against faceprints stored on the user’s phone—a database that’s currently configured to receive updates from Meta. Recognized faces will trigger notifications, while the rest are cropped, indexed, and saved to a folder marked “pending.” 尽管尚未启用，但 NameTag 存在于一款下载量超过 5000 万次的 Meta AI 配套应用中，该应用对于使用其智能眼镜（包括雷朋和欧克利型号）的关键功能至关重要。如果激活，它会将 Meta 眼镜捕捉到的人脸转换为独特的生物识别特征（通常称为“人脸指纹”），并将其与存储在用户手机上的指纹进行比对——该数据库目前配置为可接收来自 Meta 的更新。被识别的人脸会触发通知，而其余人脸则会被裁剪、索引并保存到标记为“待处理”的文件夹中。
NameTag would revive a type of technology Meta said it had sunsetted in 2021, when the company announced it would delete more than a billion faceprints belonging to Facebook users following years of controversy over its photo-tagging system. Meta ultimately paid $650 million to settle a class-action lawsuit brought by Illinois users and, in 2024, agreed to a separate $1.4 billion settlement with Texas over allegations it had unlawfully collected biometric data from users. NameTag 将复活一种 Meta 曾宣称在 2021 年已停止使用的技术。当时，在照片标记系统引发多年争议后，该公司宣布将删除超过 10 亿个 Facebook 用户的人脸指纹。Meta 最终支付了 6.5 亿美元以和解伊利诺伊州用户提起的集体诉讼，并于 2024 年就非法收集用户生物识别数据的指控，与德克萨斯州达成了另一项 14 亿美元的和解协议。
Its renewed efforts arrive amid mounting opposition to consumer-level face recognition, which privacy advocates argue will give anyone from stalkers to immigration agents easy access to a dangerous technology. Internal Meta documents published by The New York Times in February showed the company had planned to roll out the feature during a “dynamic political environment,” when Meta believed its biggest critics would be preoccupied. 在消费者层面的人脸识别技术面临日益高涨的反对声浪之际，Meta 的这一新举措随之而来。隐私倡导者认为，这将使从跟踪者到移民局官员的任何人都能轻易获取这种危险技术。《纽约时报》2 月份披露的 Meta 内部文件显示，该公司曾计划在“动态的政治环境”下推出该功能，因为 Meta 认为届时其最大的批评者将无暇顾及此事。
Three AI models powering NameTag have already been deployed from Meta’s servers and now reside on its customers’ phones, according to WIRED’s analysis, which was independently reproduced by outside experts. One model detects faces, one crops them, and a third encodes them into biometric data. 根据《连线》的分析（并经外部专家独立复现），驱动 NameTag 的三个 AI 模型已经从 Meta 的服务器部署并驻留在客户的手机中。其中一个模型负责检测人脸，一个负责裁剪，第三个则将其编码为生物识别数据。
Only traces of the user interface are currently present, hinting at how the feature may ultimately work. A May version of the app rebrands the feature for users as “Connections,” inviting them to “remember the people you met.” It remains unclear whose faces will be included in the system’s recognition database, how those profiles are created, or how many people could ultimately be identifiable through it. 目前仅能看到用户界面的痕迹，这暗示了该功能最终的运作方式。5 月份的一个应用版本将该功能重新命名为“Connections”（连接），邀请用户“记住你遇到的人”。目前尚不清楚哪些人的面孔会被纳入系统的识别数据库，这些档案是如何创建的，或者最终有多少人可能通过该系统被识别。
WIRED shared its findings with two outside security researchers who separately examined the app and reproduced key aspects of the analysis: Cooper Quintin, a security researcher and senior public interest technologist with the nonprofit Electronic Frontier Foundation’s Threat Lab, and an independent security and privacy researcher who goes by the pseudonym Buchodi and has spent more than a decade reverse engineering consumer software and surveillance technologies. 《连线》与两名外部安全研究人员分享了调查结果，他们分别检查了该应用并复现了分析的关键部分：电子前沿基金会（EFF）威胁实验室的安全研究员兼高级公共利益技术专家 Cooper Quintin，以及一位化名为 Buchodi、拥有十多年消费软件和监控技术逆向工程经验的独立安全与隐私研究员。
“The feature is not yet exposed to consumers but seems nearly ready to go,” says Quintin. “Despite the billions of reasons not to, Meta seems to have created the capacity to turn their customers into a distributed surveillance machine.” “该功能尚未向消费者开放，但似乎已准备就绪，”Quintin 表示，“尽管有数十亿个理由不这样做，但 Meta 似乎已经创造了一种能力，可以将他们的客户变成一台分布式监控机器。”
Buchodi ran additional tests on the recognition pipeline. To see if the matching system worked, Buchodi added a single faceprint to the app’s gallery, taken from the deceased French philosopher Michel Foucault. After triggering NameTag with Foucault’s image, the app produced a notification: “Person recognized.” Buchodi 对识别流程进行了额外测试。为了验证匹配系统是否有效，Buchodi 将已故法国哲学家米歇尔·福柯（Michel Foucault）的一张人脸指纹添加到了应用的图库中。在使用福柯的图像触发 NameTag 后，应用弹出了通知：“已识别出此人。”
“The main components of a face-recognition feature are already in Meta’s companion app,” Buchodi says. “Not many pieces stand between this and a working feature.” “人脸识别功能的主要组件已经存在于 Meta 的配套应用中，”Buchodi 说，“距离实现完整功能，只差临门一脚。”
In April, more than 70 advocacy groups—including the American Civil Liberties Union, Electronic Privacy Information Center, and Fight for the Future—demanded Meta scrap NameTag, warning it would let stalkers and abusers silently identify strangers in public. “Our competitors offer this type of face-recognition product, we do not,” a Meta spokesperson said in a statement to WIRED at the time. “If we were to release such a feature, we would take a very thoughtful approach before rolling anything out.” 今年 4 月，包括美国公民自由联盟（ACLU）、电子隐私信息中心（EPIC）和“为未来而战”（Fight for the Future）在内的 70 多个倡导组织要求 Meta 取消 NameTag，警告称这将使跟踪者和施暴者能够在公共场合静默识别陌生人。“我们的竞争对手提供此类人脸识别产品，我们没有，”Meta 发言人在当时给《连线》的一份声明中表示，“如果我们发布此类功能，在推出之前，我们会采取非常审慎的方法。”
Privacy advocates argue that by embedding face recognition into a mass-market wearable platform, Meta could normalize a capability it previously pulled back amid privacy concerns. 隐私倡导者认为，通过将人脸识别嵌入大众市场的可穿戴平台，Meta 可能会使一种其此前因隐私担忧而撤回的功能变得常态化。
“You’re setting norms and standards by putting technology into the ecosystem,” Joseph Jerome, a former Meta Reality Labs policy official who worked on privacy reviews for the company’s AR and VR products, says of Meta’s role in the wearable tech industry. “I don’t know how Meta can responsibly deploy a technology like this.” “当你将技术引入生态系统时，你就在设定规范和标准，”曾负责 Meta Reality Labs AR 和 VR 产品隐私审查的前政策官员 Joseph Jerome 在谈到 Meta 在可穿戴技术行业的作用时表示，“我不知道 Meta 如何能负责任地部署这样的技术。”
“Regardless of any sensational reporting, the facts are simple: We’ve said before we’re exploring these types of features, and what you’re seeing is just evidence of that exploration,” says Meta spokesperson Ryan Daniels. “Nothing has shipped to consumers and no final decision has been made on what to do here, if anything. If we do decide to roll something out, we will take a thoughtful approach.” “无论任何耸人听闻的报道如何，事实很简单：我们之前说过我们正在探索这类功能，你所看到的只是这种探索的证据，”Meta 发言人 Ryan Daniels 表示，“没有任何东西推送给消费者，也没有就此做出最终决定。如果我们决定推出某些功能，我们会采取审慎的方法。”