Transport for London's AI camera trial

Transport for London has recently piloted an AI-powered camera scheme that uses software to detect fare dodgers, potential at-risk people, disabled passengers and aggressive behaviour, using its existing CCTV equipment and image recognition software.

The technology is very impressive, and currently avoids using facial recognition - except to manually flag fare dodgers so staff can flag them to inspectors. TfL says it will improve enforcement, and could actually be useful for passengers in the long-term because staff can be made aware of issues on the platforms and elsewhere in the station that they might not see otherwise.

But is this just useful tech - or are there civil liberty and privacy concerns TfL needs to contend with? I'd be grateful for thoughts from tech and civil rights experts on what issues this could present - and the benefits it could offer.

https://takes.jamesomalley.co.uk/p/tfls-ai-tube-station-experiment-is