GPS is a one-way system, how could that ever be proof? You'd need to send people devices with some DRM on it so that they can't modify the code it runs and the check it performs (we all know how well DRM works anyway, or how desirable it is)
But I, as the attacker, would just modify the value right?
It's not that the device transmits signals into space and the satellite operator, a trusted third party, would relay to the server where the user was computed to be. Instead, it's the user self-reporting the computed value from GPS satellites' signals
If you decompiled the app, then yes, you could spoof GPS. Still, a well behaved backend would stop you in your tracks.
The user is not self-reporting, the app is.
Again, just because something can't be 100% bulletproof it doesn't mean it needs to stay wide open.
That's a fair argument in general, though in this case I would both say that
- the risk of somehow abusing self-reported city-level location data is very low
- the effort involved in bypassing the proposed security measure is so exceedingly little, anyone who is passingly motivated to abuse it will also simply do that
As someone who works in the digital risk business, I fully subscribe to managing risk and that (as you say) 100% security is usually not a realistic option. Weighing risks against benefits is key, though
I'm less concerned about the abuse of a single self-reported city-level location data. What worries me here is a world-wide-open directory of people (any people) and their locations and interests and social links. It's a target group right there served on a silver platter to _anyone_ on the internet. Sorry, but, no thanks.
GPS could confirm that you are indeed where you claim you are. In the unlikely case you want to change your location for the Nth time.