多数美国人不信任使用面部识别技术的公司

一项新的调查显示，多数美国人支持执法部门使用面部识别技术，但对企业使用这种技术感到不安。

皮尤研究中心周四（9月5日）公布研究结果称，56%受访者表示其相信执法部门会负责任地使用面部识别技术。与此同时，59%受访者表示，可以接受警察在公共场合使用这种技术评估安全威胁。

但是，美国人对使用面部识别软件的公司持怀疑立场。只有36%受访者相信科技公司会负责任地使用这种技术，而称其相信广告商的人只有17%。

调查结果显示，就尖端技术而言，美国人对执法部门和企业的信任存在巨大差别。这种态度可能会对面部扫描技术未来的监管产生重大影响。

批评人士越来越多地发出警告称，面部识别技术会对隐私产生影响，比如说这项技术能在公共场合识别守法公民。他们还担心，警方可能会利用这种技术对少数族裔进行不恰当的侧写。

皮尤研究中心数据实验室主任阿伦·史密斯解释道，调查结果之所以会凸显出这种分歧，是因为美国人愿意“放弃一部分隐私权和公民自由权”，但前提是他们觉得这样可以更好地保护其免受恐怖袭击或暴力犯罪。

史密斯说，“人们（对面部识别技术）的看法与其对执法部门以及更广泛执法的信任度密切相关。”。

虽然面部识别技术已存在多年，但直到最近才得到更加广泛的应用。得益于人工智能领域中的深度学习技术的创新，使得面部识别技术可以更准确地识别人脸。

过去几年时间里，苹果等科技巨头已经把面部识别技术作为解锁智能手机的一种安全措施， Facebook则依靠这种技术来自动识别照片中的人。与此同时，亚马逊一直试图将此技术出售给执法部门,以打击犯罪行为。

然而，公众对科技公司的看法正在滑坡，对其保护用户数据的信任度也在下降。近年来，多家科技巨头都在隐私方面犯下大错，其中包括Facebook，该公司最终支付50亿美元与美国联邦贸易委员会就涉及政治咨询公司剑桥分析的安全松懈问题达成了和解。

皮尤在6月份进行了此次调查，选取了4000多名具有代表性的美国成年人作为样本。该调查是皮尤研究中心一项大规模研究的一部分内容，这项研究旨在分析公众对数字隐私的看法。

此外，调查还考察了美国人如何看待面部识别技术准确性问题。总体而言，调查发现大多数人对这项技术的能力抱有乐观看法。

73%的美国成年人“认为面部识别技术至少能在一定程度上有效地准确识别个体”。与此同时，63%的人相信这项技术可以准确地区分性别，认为该技术可以识别种族的人占61%。

事实上，许多研究都表明面部识别技术对白人男性的识别度高于其他群体（包括女性和非裔美国人）。今年8月，美国公民自由联盟表示，面部识别技术在一次测试中将26名加州议员与“拘捕照片数据库”中的照片进行了错误匹配，而在被误认的政客中，超过一半是“有色人种议员”。

“我们发现，对这项技术听得越多的人就越是相信其效果。” 史密斯说。“我想说，人们对这种技术的总体认知相当广泛，但相对肤浅。”

与一般人群相比，非裔美国人对执法部门使用面部识别技术的怀疑态度更深。只有41%的黑人受访者称其信任执法部门使用这种技术，而在白人当中，这一比例为61%。

调查发现，与年长的美国人和共和党人相比，年轻人和民主党人对执法部门使用面部识别技术的信任度也比较低。（财富中文网）

译者：艾伦

审校：夏林

Most Americans favor law enforcement using facial recognition technology, but they are uncomfortable with corporations using it, according to a new survey.

Fifty-six percent of respondents said they trust law enforcement to use facial-recognition technology responsibly, the Pew Research Center found in a study published on Thursday. Meanwhile, 59% said it was acceptable for police to use the technology to assess security threats in public.

But Americans are far more skeptical about companies using facial recognition software. Only 36% said they trust tech companies to use it responsibly while a minuscule 17% said they trust advertisers.

The findings show a big divide among Americans when it comes to who they trust with the cutting-edge technology. That attitude could have a big impact on any future regulation of the facial-scanning technology.

Critics are increasingly raising the alarm about its privacy implications, including its ability to identify law-abiding people in public. They also fear that police could use the technology to improperly profile minorities.

Aaron Smith, director of Data Labs for the Pew Research Center, explained the divide highlighted by survey's result by saying that Americans are willing to “give up elements of their privacy and civil liberty” if they feel it would better protect them from terrorist attacks or violent crime.

“People’s perceptions [about facial recognition] map closely to their views about law enforcement and their trust in law enforcement more broadly,” Smith said.

Although facial recognition technology has existed for years, it has recently gained wider use. Because of recent innovations in deep learning, a subset of artificial intelligence, it can more accurately identify people.

Over the past couple of years, tech giants like Apple have adopted the technology as a security measure for its customers to unlock their smartphones while Facebook relies on it to automatically identify people in photos. At the same time, Amazon has been trying to sell the technology to law enforcement for fighting crime.

However, the public perception of tech companies is declining along with trust in their ability to safeguard user data. In recent years, a number of tech giants have had significant privacy blunders including Facebook, which ended up paying $5 billion to settle accusations by the Federal Trade Commission about its lax security involving political consulting firm Cambridge Analytica.

Pew's survey, conducted in June, was based on responses from a representative sample of over 4,000 adult Americans. It was part of a larger Pew Research study analyzing public perception about digital privacy.

Additionally, the survey looked at how accurate Americans think facial recognition technology is. In general, it found that most have a rosy view of the technology’s capabilities.

Seventy-three percent of U.S. adults “think facial recognition technologies are at least somewhat effective at accurately identifying individual people.” Meanwhile, sixty-three percent said they believe the technology can accurately classify a person’s gender while 61% think it can identify race.

In fact, a number of studies have shown that facial recognition technology works better on white males than others including women and African-Americans. In August, the American Civil Liberties Union said facial-recognition technology incorrectly matched 26 California lawmakers with images from an "arrest photo database" during a test, with more than half of the misidentified politicians being "lawmakers of color."

“One of the things we found is that people who have heard more about the technology have higher views of the efficacies,” Smith said. “I would say people’s awareness of the technology overall is fairly broad but relatively shallow.”

In contrast to the general population, African Americans are more skeptical of law enforcement using facial recognition technology. Only forty-one percent of black respondents said they trust law enforcement with the technology compared to 61% of whites.

Other respondents, like younger adults and those who identified as Democrats, also expressed less trust in law enforcement’s use of facial recognition compared to older Americans and Republicans, the survey found.