The other thing it knows is that you have provided a proof that you are a member of a the group conversation associated with the group call (basically a proof that you are allowed to join the call). And that proof is based on Signal's "zkgroup" system.
That's it. Once the server has the proof, it forwards packets for you. And it doesn't learning about you from then proof (other than what the proof proves: that you can join).
But isn't knowing the IP a decently good identifier? I know signal is trying to be as trustless as possible, but since IPs don't change often can't this be used to deanonymize people relatively easily? I mean supposing the server became hostile/hijacked?
This was already the case before group video calls and, yes, IPs can be used as identifiers. I don't think this can be prevented on a technical level[0] – you will always need to trust the server here. Signal does have a good track record, though: https://signal.org/bigbrother/
[0]: Unless you do p2p routing of video calls – with the known downsides regarding traffic, performance and NAT. But even then you would probably still need a server as rendez-vous point.
The other thing it knows is that you have provided a proof that you are a member of a the group conversation associated with the group call (basically a proof that you are allowed to join the call). And that proof is based on Signal's "zkgroup" system.
That's it. Once the server has the proof, it forwards packets for you. And it doesn't learning about you from then proof (other than what the proof proves: that you can join).