ARTHUR GOLDSTUCK: Behavioural data and the see-all way to catch a thief
The answer, clearly, is to identify potential criminals without identifying individuals
Sometimes new technology is too good for its own good. That was the case for Amazon Web Services two years ago, when it unveiled Rekognition Video. Described as "a deep-learning powered video analysis service that tracks people, detects activities, and recognises objects, celebrities, and inappropriate content", it was expected to transform policing and public safety.
Then the American Civil Liberties Union and MIT showed that Rekognition had an excessive level of false positives in identifying people of colour. Ironically, however, its biggest challenge is the fear that it will work too well: when anyone can be identified by surveillance systems, no-one is safe from the system being abused...
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.
Subscribe now to unlock this article.
Support BusinessLIVE’s award-winning journalism for R129 per month (digital access only).
There’s never been a more important time to support independent journalism in SA. Our subscription packages now offer an ad-free experience for readers.
Cancel anytime.
Questions? Email helpdesk@businesslive.co.za or call 0860 52 52 00. Got a subscription voucher? Redeem it now.