内容来源：https://www.wired.com/story/wrongful-arrest-tests-one-of-the-oldest-police-face-recognition-tools-in-the-us/

内容总结：

佛罗里达州男子因警方错误使用人脸识别技术遭误捕，起诉索赔

据本周三提交的一份诉讼文件，佛罗里达州一名男子因警方依赖不准确的人脸识别匹配结果而被错误逮捕，指控其企图非法引诱儿童。尽管该男子实际居住地距离案发地超过300英里（约480公里），且声称从未踏足案发城市。

52岁的罗伯特·迪伦来自迈尔斯堡，是一名商业捕蟹人。他因佛罗里达州皮内拉斯县警长办公室运营的“FACES”人脸识别系统将其面部与一张手机拍摄的电脑屏幕照片匹配，而被警方逮捕。调查记录显示，该系统给出的面部特征匹配度为“93%”。诉讼指出，这一分数仅代表算法认为两张图像的相似程度，而非两者为同一人的概率。

FACES系统存储了佛罗里达州数千万张嫌犯照片和驾照照片，是美国运行时间最长的警方人脸识别数据库之一。

美国公民自由联盟（ACLU）作为原告提起诉讼，称迪伦是在家中当着妻子的面被捕，随后在冰冷的牢房里关押了一夜，并被用装有笼子、没有照明的货车转运。他不得不抵押卡车所有权才得以保释。被捕时正值石蟹捕捞旺季，这导致他拖欠房租，险些失去住所。他的嫌犯照片在网络留存近一年，直到一名电视台记者介入后才从县网站上删除。

诉讼称，如今有陌生人在公共场合接近迪伦询问此案，他也再不敢与儿童交谈。

案件发生在2023年11月2日午夜前，地点是杰克逊维尔海滩的一家麦当劳。一名男子涉嫌接近一名12岁以下女孩，多次要求她跟自己走。女孩拒绝后，该男子第二次靠近她，女孩随即呼叫母亲。男子在警察赶到前离开。

诉讼列出了多项证据显示迪伦并非嫌疑人，但这些信息并未提交给签署逮捕令的法官。麦当劳经理告诉调查人员，该嫌疑人是“常客”，曾多次到店。而迪伦从未去过杰克逊维尔海滩，住址距离当地数百英里。

办案警官曾于当月向周边机构发送人脸识别协查通报。杰克逊维尔警局一名警长通过FACES系统比对后，给出了迪伦名字的“93%匹配”结果。调查警官随后申请查询案发前后迪伦名下两辆车的车牌识别记录，但结果显示这些车辆从未在该县出现——诉讼称，这些信息被从逮捕令申请材料中删去。

此后六个月，案件毫无进展。直到2024年7月，警官才提交逮捕令，法官签字，迪伦于次月被捕。他聘请了刑事辩护律师，并在10月做无罪抗辩。州检察官办公室在几周后撤销了所有指控。然而，那名调查警官仍在年底获得晋升。

“我永远无法忘记当时的恐惧和担忧，不知道自己还能否回家见到妻子和女儿。”迪伦在一份声明中说，“一年多过去了，我仍在收拾生活的残局，这一切都是因为警方依赖这种危险的技术，而不是尽职调查。”

诉讼将调查警官和杰克逊维尔警局警长列为被告，同时以职务身份起诉了杰克逊维尔海滩市、杰克逊维尔警长和皮内拉斯县警长。原告寻求补偿性和惩罚性赔偿，并要求法院责令上述三个机构全面整改其人脸识别政策。

杰克逊维尔警局发言人在一份声明中表示：“由于诉讼正在进行，我们无法进一步评论此事。”皮内拉斯县警长办公室则未立即回复置评请求。

杰克逊维尔警长T.K.沃特斯在案件撤销后接受当地媒体采访时表示，仅凭人脸识别结果不能构成其部门的逮捕依据。“如果有人拿着人脸识别匹配结果来找我，这就是你的全部证据，我可能会把你赶出办公室。”

FACES系统自2001年起由皮内拉斯县警长办公室运营，是美国最早的警方人脸识别系统之一。2021年高峰期，其数千万张佛罗里达州嫌犯及驾照照片可供包括FBI和移民执法局在内的260多个机构查询。调查人员上传嫌疑人图像后，系统会与数据库比对并返回排序列表。

该系统长期缺乏监管。2016年乔治城大学法学院隐私与技术研究中心研究发现，皮内拉斯县警长办公室从未审查数据库检索记录，也不需要合理怀疑即可发起查询。当被问及是否审计检索滥用情况时，时任警长鲍勃·瓜尔蒂耶里回答：“不，基本没有。”据《太阳哨兵报》和普利策中心报道，佛罗里达州机构还曾用FACES系统扫描和平抗议者的面部。

ACLU指出，迪伦案只是美国已知至少15起由人脸识别技术导致的冤假错案之一。今年早些时候，同一家杰克逊维尔警局在汽车盗窃调查中误捕了一名北卡罗来纳州男子。据当地媒体报道，他因85%的匹配率被捕，在监狱关押近三个月，等到指控撤销时，已经失去了房子、工作以及两个孩子的监护权。

“没有人应该因为算法出错而失去自由，或者不敢出门。”ACLU言论、隐私和技术项目副主任内特·韦斯勒呼吁佛罗里达州警方纠正错误，并制定保护措施防止未来再发生冤案。他补充道：“全美警察都该警醒了。不可靠的人脸识别技术正在伤害民众，我们将继续追究这些滥用行为的责任。”

中文翻译：

佛罗里达州一名男子近日提起诉讼称，警方依赖不准确的面部识别匹配结果，错误地逮捕了他，指控其试图非法引诱儿童——尽管他住在距事发地300多英里（约483公里）之外，且声称从未踏足犯罪发生的城市。这起诉讼于周三提交。

52岁的罗伯特·狄龙来自迈尔斯堡，是一名商业捕蟹人。他因FACES系统（由佛罗里达州皮内拉斯县警长办公室运营的面部识别系统）将其面部与一张用手机拍摄的电脑屏幕上的男子照片匹配成功而被捕。警方调查记录显示，系统给出的“面部特征匹配度为93%”。系统输出的分数代表两幅图像在算法眼中的相似程度，而非它们显示的是同一人的概率。

FACES系统储存了佛罗里达州数千万张嫌犯照片和驾照照片，是美国运行时间最长的警方面部识别数据库之一。

提起此诉讼的美国公民自由联盟表示，狄龙是在家中当着妻子的面被捕，随后被关押在冰冷的牢房里过夜，并被一辆装有笼子、没有灯光的货车转运。他为获得保释抵押了卡车所有权。被捕正值石蟹捕捞旺季，导致他拖欠房租，险些失去家园。他的嫌犯照片在网上留存近一年，直到一名电视记者介入后才被从县网站上撤下。

诉状称，常有陌生人在公共场合找狄龙询问此案，他现在已不愿与儿童交谈。

事件发生在2023年11月2日午夜前不久，地点是杰克逊维尔海滩的一家麦当劳。一名男子涉嫌接近一名12岁以下女孩，多次要求她跟自己离开。女孩拒绝后，该男子再次接近她，女孩随即呼唤母亲。该男子在警方到达前离开。

诉状列出了多项与狄龙无关的事实，而这些信息从未呈递给签署逮捕令的法官。麦当劳的一名经理告诉调查人员，该嫌疑人是她见过多次的“常客”。诉状称，狄龙从未去过杰克逊维尔海滩，他住在数百英里之外。

同年11月，杰克逊维尔海滩一名负责此案的警员使用手机拍摄的麦当劳监控录像画面，向周边机构发送了身份识别公告。杰克逊维尔警长办公室的一名警官通过FACES系统对图像进行比对，并传回“93%匹配”结果，关联到狄龙的名字。调查警员随后申请查询车牌识别系统，寻找狄龙名下两辆车在案发前后的行踪。诉状称，两辆车均未在县内任何地点出现，而这一结果在逮捕令申请中被遗漏。

诉状指出，此后的六个月没有进一步调查。2024年7月，该警员提交逮捕令申请。法官签署逮捕令，狄龙于次月被捕。他聘请刑事辩护律师，并于同年10月提出无罪抗辩。几周后，州检察官办公室撤销所有指控。然而，该调查警员仍在年底获得晋升。

狄龙通过律师发布声明表示：“我永远无法忘记当时的恐惧与担忧，不知道自己还能否回家与妻女团聚。一年多过去了，我仍在收拾生活残局，而这一切只因警方依赖这种危险的技术，却没有尽职尽责地展开真正调查。”

诉讼将调查警员和杰克逊维尔警长办公室的警官列为个人被告，同时以职务身份起诉杰克逊维尔海滩市、杰克逊维尔警长及皮内拉斯县警长。诉讼要求赔偿性及惩罚性赔偿，并请求法院命令三家机构全面改革其面部识别政策。

杰克逊维尔警长办公室发言人在给《连线》杂志的声明中表示：“由于诉讼正在进行，我们无法就此事进一步评论。”皮内拉斯县警长办公室未立即回应置评请求。

杰克逊维尔警长T.K.沃特斯在案件撤销后告诉当地新闻台Action News Jax，仅凭面部识别匹配结果不能构成其办公室的逮捕依据：“如果你拿着面部识别匹配结果来找我，并以此作为可能原因，我可能会把你赶出办公室。”

FACES系统自2001年起由皮内拉斯县警长办公室运营，是美国历史最悠久的警方面部识别系统之一。2021年其巅峰时期，超过260个机构（包括联邦调查局和移民和海关执法局）可调取系统内数千万张佛罗里达州嫌犯照片和驾照照片。使用时，调查人员上传嫌疑人图像，系统与数据库比对后返回排名列表。

该系统长期以来几乎不受监管。乔治城法律中心隐私与技术项目2016年研究显示，皮内拉斯县警长办公室未对数据库搜索进行审计，且无需合理怀疑即可发起查询。当被问及是否审计滥用搜索行为时，警长鲍勃·瓜尔蒂耶里回答：“不，基本没有。”据《太阳哨兵报》和普利策中心报道，佛罗里达州机构还曾使用FACES系统扫描和平抗议者。

美国公民自由联盟称，狄龙案是美国已知至少15起因面部识别技术导致的错误逮捕案件之一。今年早些时候，同属杰克逊维尔警长办公室在另一起汽车盗窃调查中错误逮捕了一名北卡罗来纳州男子。据Action News Jax报道，该男子因85%匹配度被捕，在监狱关押近三个月。指控撤销时，他已失去住房、工作和两名子女的监护权。

美国公民自由联盟言论、隐私与技术项目副主任内特·韦斯勒表示：“没有人应该因为算法出错而失去自由或不敢出门。”他呼吁佛罗里达州警方进行补救，并采取保障措施防止未来发生类似错误逮捕。“全国警方都该警醒了，”韦斯勒补充道，“不可靠的面部识别技术正在伤害民众，我们将继续追究他们的责任，制止这类滥用行为。”

英文来源：

A Florida man was wrongfully arrested for attempting to illegally lure a child after police relied on a face-recognition match that was inaccurate, according to a lawsuit filed on Wednesday, even though he lived more than 300 miles from the scene and says he had never set foot in the city where the crime took place.
Robert Dillon, a 52-year-old commercial crabber from Fort Myers, was arrested after FACES—a face-recognition system operated by Florida’s Pinellas County Sheriff's Office—matched his face against a photo of a man on a computer screen taken with a cell phone. The system returned a “93 percent match on facial features,” according to police-investigatory notes. The scores it emits represent how much two images look alike to the algorithm. Not how likely it is that they show the same person.
FACES holds tens of millions of Florida mug shots and driver's license photos and is one of the longest-running police face-recognition databases in the United States.
The American Civil Liberties Union, which filed the suit, says Dillon was arrested at his home in front of his wife, held overnight in a cold cell, and transported in a caged, unlit van. He pledged the title to his truck to make bond. The arrest came during peak stone crab season, causing him to fall behind on rent and nearly lose his home. His mug shot stayed online for nearly a year, removed from the county website only after a TV reporter intervened.
Strangers approach Dillon in public to ask about the case, the complaint says, and he no longer feels comfortable talking to children.
The incident took place shortly before midnight on November 2, 2023, at a McDonald's in Jacksonville Beach, where a man allegedly approached a girl under 12 and repeatedly asked her to leave with him. She refused. After he approached her a second time, she called for her mother. The man left before the police arrived.
The complaint lays out several facts that pointed away from Dillon and never reached the judge who signed the warrant for his arrest. A manager at the McDonald's told investigators the suspect was a “regular customer” she had seen there multiple times. According to the complaint, Dillon had never visited Jacksonville Beach, living hundreds of miles away.
A Jacksonville Beach police officer assigned to the case sent an attempt-to-identify bulletin to surrounding agencies later that November using cell phone photos of the McDonald’s surveillance footage. A sergeant with the Jacksonville Sheriff's Office (JSO) ran the images through FACES and sent back the “93 percent match” to Dillon's name. The investigating officer then requested a search of license plate readers for two vehicles registered to Dillon, covering the days around the incident. Neither turned up anywhere in the county, according to the complaint, which says the results were omitted from the warrant application.
Six months passed with no further investigation, the complaint says. In July 2024, the officer submitted the warrant. A judge signed it, and Dillon was arrested the following month. He retained a criminal defense attorney and, that October, pleaded not guilty. The State Attorney's Office dropped all charges a few weeks later. The investigating officer was nevertheless promoted by the end of the year.
“I will never get over how terrified and worried I was, wondering if I’d ever go home to my wife and daughter again,” Dillon says in a statement shared by his attorneys. “Over a year later, I'm still picking up the pieces of my life, all because the police relied on this dangerous technology instead of doing their jobs and actually investigating.”
The suit names the investigating officer and the JSO sergeant individually, and also targets the City of Jacksonville Beach, the Jacksonville Sheriff, and the Pinellas County Sheriff in their official capacities. It seeks compensatory and punitive damages, and asks a court to order all three agencies to overhaul their face-recognition policies.
“Due to pending litigation, we would be unable to comment further on the incident,” a Jacksonville Sheriff's Office spokesperson tells WIRED in a statement. The Pinellas County Sheriff's Office did not immediately respond to a request for comment.
Jacksonville Sheriff T.K. Waters told local news station Action News Jax after the case was dropped that a face-recognition hit alone would not constitute probable cause in his office: “If you came to me with a facial-recognition hit and that was your probable cause, I would probably kick you out of my office,” he said.
FACES has been operated by the Pinellas County Sheriff's Office since 2001, making it one of the oldest police face-recognition systems in the country. At its peak in 2021, its tens of millions of Florida mug shots and driver's license photos were accessible to more than 260 agencies—including the FBI and ICE. To use the system, investigators upload an image of a suspect, the system compares it against the gallery, and it returns a ranked list of possible matches.
The system has long operated with little oversight. A 2016 study by Georgetown Law's Center on Privacy and Technology found that Pinellas County Sheriff's Office conducted no audits of how the database was searched and required no reasonable suspicion to run a query. Asked whether the office audited searches for misuse, Sheriff Bob Gualtieri replied, “No, not really.” Florida agencies have also used FACES to scan peaceful protesters, according to reporting by the Sun Sentinel and Pulitzer Center.
ACLU says Dillon's case is one of at least 15 known wrongful arrests in the United States attributed to face-recognition technology. Earlier this year, the same Jacksonville Sheriff's Office wrongfully arrested a North Carolina man in an auto-theft investigation. According to Action News Jax, he spent nearly three months in jail after an 85 percent match led to his arrest. By the time the charges were dropped, he had lost his home, his job, and custody of his two kids.
“No one should lose their freedom or be scared to leave their house because an algorithm got it wrong,” says Nate Wessler, deputy director of the ACLU's Speech, Privacy, and Technology Project, who called on Florida police departments to make amends and adopt safeguards to prevent future wrongful arrests.
“Police across the country are on notice,” adds Wessler. “Unreliable face-recognition technology is hurting people, and we will keep fighting to hold them accountable for these abuses.”