I was wrongly arrested when AI facial recognition system identified me as a thief - but this is what the REAL criminal actually looks like

✕ Sensationalism: The headline uses a first-person narrative and emphasizes mistaken identity in a dramatic way, potentially overselling the contrast for clicks.

"I was wrongly arrested when AI facial recognition system identified me as a thief - but this is what the REAL criminal actually looks like"

✕ Framing By Emphasis: The lead emphasizes the physical dissimilarity between the innocent man and the real thief, framing the error as obviously preventable, which may downplay technical or procedural nuance.

"It is hard to imagine how the pair could be mistaken for each other, with Mr Choudhury possessing a head of curly hair and Zlatineanu, 23, a shorter black style."

✕ Loaded Language: Phrases like 'misery ordeal' and 'horrific' reflect the subject’s emotional state but are presented without sufficient distancing, risking emotional sway.

"'No tech company would ever put a system into production with a failure rate of one in 25. That's horrific. It is filled with bugs.'"

✕ Editorializing: The article quotes the subject’s accusation of racial discrimination without counterpoint or neutral framing, potentially amplifying unverified claims.

"'You've probably just seen two brown people, even though they have completely different features and said, "yeah, they look close enough. Let's arrest them."'"

✕ Appeal To Emotion: Highlighting that stolen money was for flood victims in Sri Lanka adds emotional weight unrelated to the core issue of facial recognition errors.

"swipe a heap of cash which had been donated to support victims of recent floods in Sri Lanka"

✓ Proper Attribution: Key claims are attributed to named individuals and official sources, including the suspect, the wrongly accused, and police.

"TVP confirmed Mr Choudhury's arrest was 'based on the investigating officers' own visual assessment'"

✓ Comprehensive Sourcing: Includes perspectives from the victim of mistaken arrest, police confirmation, and court sentencing details, offering multiple angles.

✓ Comprehensive Sourcing: Provides background on the false arrest timeline, prior mugshot usage, and system failure rate among Asian faces, adding technical and procedural context.

"He previously said he blames the software - which returns false matches 4 per cent of the time among Asian faces - as well as the detectives analysing the clips."

✕ Omission: Fails to clarify why the real suspect, arrested December 8, was not charged until January 12, nor why the AI match triggered action weeks later despite prior arrest.

✕ Cherry Picking: Focuses heavily on physical differences but omits any explanation of how the AI system or human reviewers justified the match, limiting full context.