‘Do I look like this?’ Bangladeshi man wrongfully detained in UK after facial recognition software identified him as a thief
Alvi Choudhury, a Bangladeshi-born software engineer, was arrested at his home in Southampton last month after facial recognition software mistakenly identified him as a burglary suspect in Milton Keynes, 100 miles away, and he was detained for 10 hours. Chaudhry told the Guardian he was confused when police knocked on his door and arrested him at work. When he saw the CCTV footage of the break-in, he was furious because the killer looked nothing like him. “I was very angry because this kid looked about 10 years younger than me,” said the bearded Chowdhury. “Everything is different. The skin is lighter. The suspect looks 18 years old. His nose is bigger. He has no hair on his face. His eyes were different. His lips are smaller than mine.“I just thought investigators saw me as a guy with curly brown hair and decided to arrest me.”Chowdhury said officers from Hampshire Constabulary laughed when he asked them “Does this look like me?”He added: “After seeing the suspect’s video and my photo, they knew I was not a suspect.”Chaudhry’s photo is in the system because he was arrested once in 2021, which he called a wrongful arrest. He said he was attacked while on an outing at a university in Portsmouth. He was released with no further developments in the case. But now he fears his picture is in the system again and if a brown man in Scotland robs a bank, they will come after him. Chaudhry is suing police for arresting him at his home, causing a scene in which neighbors watched him being led away in handcuffs.
Police say arrest was wrong but there was no racial bias
Thames Valley Police admitted the arrest was a mistake but insisted it was not due to any racial prejudice. A police spokesman told the Guardian: “While we apologize for the distress caused to the complainant in this case, their arrest was based on investigators’ own visual assessment that the individual matched the suspect on CCTV footage following a retrospective facial recognition match and was not influenced by racial profiling.”
