The US government backtracked on a plan to take photos of everyone entering the country, even US citizens, while the Chinese now plan to take photos of anyone buying a SIM card.
Meanwhile Chinese companies are trying to play a role in how the ITU sets common standards for how face recognition cameras work – this might sound trivial, but big non-tech companies will often say “we are buying X thousand cameras and they must meet ITU standard such-and-such”, so Chinese companies (and by extension their government) would be feeding their opinion into how this should work globally.
The underlying point for all of these is that machine learning means computers, or artificial intelligence can ‘see’ – instead of cameras taking photos that a human somewhere has to look at, they can produce a stream of searchable, indexable data that a computer can use like any other data input.
