由众议员史蒂文·霍斯福德（民主党，内华达州）和参议员伊丽莎白·沃伦（民主党，马萨诸塞州）领导的一个由数十名跨党派议员组成的团体，对谷歌计划获取“精神航空”（Spirit Airlines）的数据以训练其人工智能模型表示担忧。
A bipartisan group of dozens of lawmakers led by Rep. Steven Horsford (D-Nev.) and Sen. Elizabeth Warren (D-Mass.) is concerned over Google’s plan to acquire Spirit Airlines data to train its AI models.
该团体致信谷歌首席执行官桑达尔·皮查伊和“精神航空”首席执行官戴夫·戴维斯，对拟议的“以训练人工智能系统为目的，将‘精神航空’的内部数据出售给谷歌”的1000万美元交易提出质疑。议员们认为，此举可能侵犯“数千名‘精神航空’现任及前任员工的隐私和机密信息”。
A letter from the group, sent to Google CEO Sundar Pichai and Spirit CEO Dave Davis, took issue with a proposed $10 million sale of “Spirit Airlines’ internal data to Google for the purpose of training artificial intelligence systems,” which the lawmakers said they believed could violate “the privacy and confidentiality of thousands of current and former Spirit employees.”
该团体由来自两院的119名议员组成，他们指出，此次交易属于“前所未有的领域”，因为涉及“约1亿封电子邮件、5亿条Microsoft Teams消息、员工档案、工时记录、薪资和税务信息以及雇佣合同”。议员们在信中写道：“我们担心，在当代人工智能的背景下，传统的去标识化保护措施可能不足以保障员工隐私。”他们援引了美国国家标准与技术研究院的一份报告，该报告指出，在某些情况下，去标识化的数据仍可能通过人工智能被重新识别。
The group of 119 lawmakers from both chambers continued the sale is “uncharted territory” as it would involve “approximately 100 million emails, 500 million Microsoft Teams messages, employee records, timecard records, payroll and tax information, and employment contracts.”“We are concerned that conventional de-identification safeguards may not be sufficient to protect employee privacy in the context of modern artificial intelligence,” the lawmakers wrote, citing a National Institute of Standards and Technology report that stated in certain circumstances, de-identified data could be re-identified with the use of AI.
据一位了解此事的消息人士称，谷歌已收到这封信，且无意接收任何个人信息，并补充说，任何收到的数据都将由独立的第三方进行严格去标识化处理。
Google received the letter and has no interest in receiving personal information, according to a source familiar with the matter, adding that any data received will be rigorously de-identified by an independent third party.
霍斯福德在周四的一份新闻稿中表示：“创新不能以牺牲工人隐私为代价，在真正具有约束力的保障措施到位之前，任何员工数据都不应被转移。”
“Innovation cannot come at the expense of workers’ privacy, and no employee data should move until real, enforceable safeguards are in place,” Horsford said in a press release on Thursday.
谷歌发言人告诉《国会山报》：“正如我们从这一流程一开始就所说，我们并不打算从Spirit公司购买任何个人信息。”“这些信息要么会被完全排除，要么在谷歌接收任何数据之前，都会由独立的第三方进行去标识化处理。我们已经在与指定的隐私监察员进行建设性的合作，”该发言人继续说道。
A Google spokesperson told The Hill: “As we’ve said from the beginning of the process, we are not looking to buy any personal information from Spirit.”“The information will either be completely excluded or will be deidentified by an independent third party before Google receives any data. We’re already working constructively with the appointed privacy ombudsman,” the spokesperson continued.
议员们要求谷歌和Spirit公司采取六项措施——包括尽可能排除员工信息、制定受影响员工同意的去标识化协议，以及在去标识化时保护员工的敏感信息。
The lawmakers asked Google and Spirit to undertake six steps — including excluding employee information as much as possible, establishing a de-identification protocol that affected employees agree to and protecting sensitive employee information when it is deidentified.
随着谷歌、OpenAI、xAI和Anthropic等大型科技公司开发人工智能模型，相关担忧日益加剧。上周，这四家公司的领导层均受邀前往白宫，并同意了一项自我监管人工智能发展的承诺。
Worry has increased surrounding the development of AI models at major technology firms such as Google, OpenAI, xAI and Anthropic. Leadership at all four companies were invited last week to the White House, where they agreed to a pledge to self-police AI development.
在承诺公布后，皮查伊在社交媒体上写道：“谷歌致力于以正确的方式构建模型，包括适当的测试、评估、红队测试以及其他防止滥用和偏离目标的安全措施——只有在模型或产品经过全面审查之后，才会发布。”
“Google’s focused on building the right way, with appropriate testing, evaluations, red-teaming, and other safeguards against misuse and misalignment – and releasing models or products only after they’ve been thoroughly reviewed,” Pichai wrote on social media after the pledge was announced.
谷歌还与五角大楼达成了一项协议，同意为其“任何合法的政府用途”提供其人工智能模型。五角大楼的一名官员在4月告诉《国会山报》。
Google also struck a deal with the Pentagon to provide its AI models for “any lawful governmental purpose,” a Pentagon official, told The Hill in April.
这项协议遭到了该公司员工的反对，他们认为，“与这项技术的密切关系意味着他们有责任指出并防止其最不符合道德且最危险的用途。”
The deal received pushback from the company’s employees who felt that their “proximity to this technology creates a responsibility to highlight and prevent its most unethical and dangerous uses.”