None of the arguments in favor of mixing in the OP is founded on moral grounds.
I disagree. There is a dilemma.
I was just plainly reacting to the arguments that mixers keep Bitcoin fungible and freezing coins would breach the users' trust. These aren't moral arguments per se, although we could attach moral implications to them, just like we could to almost everything else.
And this dilemma lies here: if the mixer uses AML tools without restrictions like most traditional exchanges do, then it no longer serves this function. The ethical mixer protocol I linked to above is an idea to solve that dilemma.
I don't think it's as simple as that. Are we just trying to please them all or at least displease them all a little but more or less equally? Are we simply skirting heavy responsibilities while being compliant but also not too much as to lose the interest of users?
I don't think ethics is simply
playing safe hitting the perfect balance. Quite the contrary, if there's such a thing as mixer ethics, I don't think the service should reserve even a tiny place for AML tools and KYC and the like, because isn't that itself a complete betrayal to the very justification of their own existence?
Unless there's more or less absolute certainty that a deposit is stolen, everything should proceed. Mere suspicions, indirect links, and the like shouldn't be grounds for rejection.
Again, how do you know this? And much more: how to know this in real time, e.g. in the 10 minutes after the coins are stolen, where many thieves will try a first mixing step? You need an AML tool - and then you enter the dilemma described above.
I was thinking of those funds that are, borrowing from your thread, "directly coming from criminal sources". If there's an AML tool needed for that, it'd be the most rudimentary of all.
As I've said, mixers function in good faith and with the presumption of regularity. It's designed to "keep bitcoins fungible", not to catch dirty funds. Having said that, if alerts and warnings came a little later than the criminal's use of the service, that's water under the bridge.
However, again, there must be certain thresholds. At least it somehow softens the impact or perhaps even discourages the use of the service by criminals.
As far as morality is concerned, I don't think there's any dilemma at all. Helping a thief get away with stolen money or his/her crime is morally wrong, plain and simple.
(...)
Maybe you're right. Maybe it's really that simple. But personally, I find the issue more nuanced than that.
By the same logic, should we, as the Bitcoin community, be OK with, or even demand that devs/miners come together to override the protocol to reverse the hack (similar to what Ethereum devs did after DAO hack)? I know both scenarios are not a perfect analogy, but the principle is similar: we let a bad thing slide because undoing it would have much worse consequences (i.e. a complete loss of trust for Bitcoin or mixers).
If it isn't black and white, and confronted with a choice between a smaller evil and a bigger evil, I'd obviously prefer the smaller evil.
These are practical matters, and I won't mind taking a utilitarian approach.
Or another angle: if it's not immoral for a person to launch a decentralised, automated mixing service (like CoinJoin), would it become immoral if the same type of service had a manual, human component to it, and the operator chose to have a "no-questions-asked" policy?
I think the automated service itself is merely a tool, just like a fork or a piece of paper. It's neither moral or immoral. It's amoral. It serves a neutral purpose. In the case of a mixer, for Bitcoin's fungibility or privacy.
If there were ulterior motives in the process of developing it, that's with the creator, not the result or product. If the knife's original inventor had killing as his/her primary goal, I guess that won't matter now. As to the product or service itself, I'd probably still lean toward the consequentialist way of assessing it. If mixers are beneficial only or mainly to criminals, then that's obviously immoral. But if everybody else is using it for their privacy, I don't find any moral issue with it.
If it's not truly automated and an operator unwittingly let a dirty fund pass through, no problem at all. If the operator chose to turn a blind eye, that's problematic. But I'm afraid we won't be talking about intentions as if they can truly be known.