Well, it serves me right to have it be my turn to answer questions!
matchboxbananasynergy How do we arrive to this conclusion? What attack are you referring to here?
This was mentioned in a post above. I hope I didn't miscommunicate. Let me try quoting it here; I hope I do it right!
DeletedUser115 Extra software increases attack surface. Also, given Play Services can receive notifications it creates an extra attack channel
matchboxbananasynergy GrapheneOS also plans to have Contact Scopes as well in order to limit what contacts an app will have access to.
This would be great! I abandoned WhatsApp, for example, on iOS because of this very issue. I have many friends and colleagues that still use it as their primary messenger, so it's complicated for me to abandon it. I've only managed to convert a subset of my contacts to Signal over the years.
matchboxbananasynergy I just want to make it clear (I think you understand this, but I want to make sure) that this mutual communication is not in any way limited to Play Services etc. It's something that all apps are fundamentally capable of doing. If you think that one of your apps that houses sensitive data might be abusing this communication channel to pass on private data that you've trusted it with, perhaps the bigger issue isn't Sandboxed Google Play, but the app you're trusting with your data in the first place. Just some food for thought!
Yes, thank you for confirming, this was my understanding that all apps within a profile can do this. Other than Spotify, the apps in my owner profile are ones that come recommended by the privacy community (e.g., Signal, Protonmail, Standard Notes, Bitwarden, etc.). Am I correct in assuming that using this category of apps limits, or even negates, the privacy risks of this potential communication channel? Since it requires mutual consent between apps, I assume that even if Spotify wanted to secretly communicate with Signal, for example, that Signal would brush off those attempts, breaking the mutuality of that channel?