From @whyrusleeping on Tue Apr 07 2015 16:31:53 GMT+0000 (UTC)
ipfs makes no claims about anonymity, the default routing system (the DHT) makes it very easy to query the ip address of a peer hosting any given block. Down the road we plan on implementing a TOR-like routing system that may provide anonymity.
Copied from original issue: https://github.com/ipfs/faq/issues/12
From @jbenet on Tue Apr 07 2015 16:41:10 GMT+0000 (UTC)
We will be enumerating all parts of the protocol where anonymity is concerned, so that different modules can be built. (note that there are strong tradeoffs, which is why ipfs choses to be modular)
Does IPNS publish/update message contain information about node that published or updated the IPNS hash?
In other words, can receiver node deduce source node?
Is there any update on this enumeration and how it intersects with the modules? Is anyone working on privacy oriented modules right now? Thanks!
1 Like
Never heard of one either. Can’t really imagine how it would be implemented into design, rather than a layer at this stage.
IMHO, as a protocol IPFS does not need privacy. I am a firm believer in privacy, but I think maybe privacy should be implemented by the projects that are working on top of IPFS, each to match the desires of itself.
I am not an expert, but I have heard from some people who are that there are layers that IPFS relies upon (e.g. DNS) that leak private data like crazy. https://dnsprivacy.org/wiki/display/DP/DNS+Privacy+-+The+Problem
This means that even if you are able to encrypt data used on top of IPFS, you have still leaked a great deal of information about who is communicating with whom.
That is why GNU has the gnu:net project to addresses these privacy concerns from the bottom up. The result is a stack with similar goals to IPFS, but where privacy is not an afterthought. https://gnunet.org/en/
I think it would be great it if IPFS and gnu:net were to have a lovechild. 
Gnu net for ipfs, well… I don’t think…
See Our Focus for 2020 and Anonymous IPFS, Filecoin or whatever protocol in fact
I think IPFS must address this one day, you cannot let this sensible subject to the responsibility of the upper apps
1 Like