Skip to Content
logologo
AI Incident Database
Open TwitterOpen RSS FeedOpen FacebookOpen LinkedInOpen GitHub
Open Menu
発見する
投稿する
  • ようこそAIIDへ
  • インシデントを発見
  • 空間ビュー
  • テーブル表示
  • リスト表示
  • 組織
  • 分類法
  • インシデントレポートを投稿
  • 投稿ランキング
  • ブログ
  • AIニュースダイジェスト
  • リスクチェックリスト
  • おまかせ表示
  • サインアップ
閉じる
発見する
投稿する
  • ようこそAIIDへ
  • インシデントを発見
  • 空間ビュー
  • テーブル表示
  • リスト表示
  • 組織
  • 分類法
  • インシデントレポートを投稿
  • 投稿ランキング
  • ブログ
  • AIニュースダイジェスト
  • リスクチェックリスト
  • おまかせ表示
  • サインアップ
閉じる

レポート 3887

関連インシデント

インシデント 6912 Report
Facewatch Reported to Have Wrongfully Flagged Home Bargains Customer as Shoplifter

Loading...
Innocent woman branded a shoplifter by facial recognition software
thetimes.co.uk · 2024

Shoppers have been warned that facial recognition software used by stores is wrongly identifying some innocent customers as thieves.

A woman said that an employee at Home Bargains, the variety store chain, accused her of being a shoplifter when she was incorrectly flagged by the ­Facewatch software.

The BBC's *Newsnight *programme ­reported that the woman, who wanted to remain ­anonymous, had her bag searched, was walked out of the shop and told that she was banned from all stores that used the ­technology.

She told the BBC: "I was just crying and ­crying the entire journey home ... I thought, 'Oh, will my life be the same? I'm going to be looked at as a shoplifter when I've never stolen.' "

The BBC reported that Facewatch, which uses a cloud-based facial recognition security system, wrote to her and ­admitted it had made an error.

The company's system is used in stores throughout the UK --- including Budgens, Sports Direct and Costcutter.

Facewatch, which is based in central London, declined to comment but said its technology helped to prevent crime and protect workers. Home Bargains also declined to comment.

The BBC also reported a case of ­mistaken identity by the Metropolitan Police using facial recognition.

Shaun Thompson, who works for the youth-advocacy group Street Fathers, said that he was incorrectly identified by police using a facial recognition system near London Bridge. Officers operating a modified van with cameras on its roof approached Thompson and told him he was a wanted man.

He was asked to give fingerprints and held for 20 minutes. Thompson said he was released only after he handed over a copy of his passport. "It felt intrusive," he said. "I was treated as guilty until proven innocent."

It is understood that the mistake may have been due to a family resemblance. The Met declined to comment.

In an interview last month to mark a year of the technology's use, Lindsey Chiswick, the force's director of ­intelligence, said that it had been the biggest breakthrough for crime detection since the advent of DNA profiling.

Times analysis shows that since its introduction live facial scanning has been used 62 times over 340 hours, ­resulting in 152 arrests --- equivalent to ­almost one arrest every two hours. During the coronation in May last year a wanted sex offender was caught and sent back to prison for breaching the terms of his release.

Another alleged offender, who was wanted by three forces, was arrested for two ­robberies, thefts and drug offences. On another occasion a man was ­arrested for indecent assault and two rapes --- one in 2017.

Chiswick said: "We know the technology is accurate so the chances that the person they stop is the person who is wanted is much higher than operating without it. I would say this is transformational. The Met's been on a journey recently in order to rebuild trust and confidence in how we police London and that's really important to us."

The Met said about one in every 33,000 people who passed its cameras were misidentified. The error count was higher, however, once someone was flagged. One in 40 alerts so far this year has been in error.

Earlier this year the government ­announced that £230 million would be spent over four years on police technology, including facial recognition vans. Chris Philp, the policing minister, said: ­"Facial recognition has the potential to completely revolutionise policing."

情報源を読む

リサーチ

  • “AIインシデント”の定義
  • “AIインシデントレスポンス”の定義
  • データベースのロードマップ
  • 関連研究
  • 全データベースのダウンロード

プロジェクトとコミュニティ

  • AIIDについて
  • コンタクトとフォロー
  • アプリと要約
  • エディタのためのガイド

インシデント

  • 全インシデントの一覧
  • フラグの立ったインシデント
  • 登録待ち一覧
  • クラスごとの表示
  • 分類法

2024 - AI Incident Database

  • 利用規約
  • プライバシーポリシー
  • Open twitterOpen githubOpen rssOpen facebookOpen linkedin
  • e1b50cd