Remember that Stack Overflow and Wikipedia are different, in that StackOverflow grants a higher degree of identity/ownership/authorship to the authors of content it hosts than does Wikipedia. That both helps and hurts Stack Overflow in terms of allowing re-use. It hurts (at least for promoting re-use), in that Stack Overflow has more stringent attribution requirements (this isn't necessarily a bad thing overall, though). It helps, in that violations are now (because of the prior condition) at least attributed directly to the correct party. If infringing content is re-used correctly, there is a name attached to it that says, in a nutshell, "This person promised us he has rights to the content, and gave us permission to redistribute it. If you have a problem, take it up with him."
Now moving on with regards to the copyright issue from my previous post raised here. It's important to remember from here on out that I am not a lawyer, and I have no legal training or experience. I'm just some guy that's followed this topic quite a lot.
I think you can make a good case that "diamond mods" are granted authority by SE Inc to act in certain specific situations on direct behalf (as agents) of SE Inc. It's worth noting here that the ability to respond to DMCA requests is explicitly excluded from what mods are empowered to do, as they cannot purge the material from the database. My belief is also that to best qualify for DMCA protection as a service provider, SE Inc and it's agents have certain obligations that they not distinguish between valid and invalid content (where "valid" means valid relative to the matters covered by the DMCA). Doing so could make it complicit in (and therefore potentially liable for) the content that exists on it's site.
This liability is not covered directly (and should be excluded) by the law, but actual case law and precedent is a funny thing. My interpretation is that, to guarantee the Section 230 safe harbors, the safest course is to play dumb entirely with regards to copyright.
To understand why, first consider the opposite: that SE Inc is allowed to choose when content may be infringing. Imagine that SE Inc notices a lot of MSDN material showing up on Stack Overflow, and starts policing the site for direct copying of Microsoft's MSDN material. However, they neglect to do so for Oracle's Java/OracleDB/MySQL documentation. There's a wedge there that Oracle, if it chose to do so, could potentially use in a legal fight to create liability for Stack Overflow with regards to infringement of Oracle copyrights, because in choosing to enforce one set of copyrights but not another SE Inc is itself making a statement — an act of speech with regards to Oracle's copyrights. I think neither Mircosoft nor Oracle would want to do this, and that Stack Overflow would likely respond to this imaginary scenario by beginning to police content for both. Stack Overflow would also likely win a contest, but Oracle probably has enough to make a long and expensive fight of it. The real problem is that there's no legal difference between Oracle and any small blog out there that may have content infringed. You would ultimately have to extend this policy to monitoring everything that came in to the site for any potential violation. This, to my view, is not what is intended by Section 230.
How, then, should a diamond mod respond to a flag for copyright violation? If the flag is directly from the copyright owner they would be unequipped to respond adequately. If it's not the copyright owner, it's not the mod's place to make a decision on the copyright owner's behalf — the mod doesn't know if permission was granted on a separate channel before posting, or if the copyright owner would prefer the infringing content remain online (for example in the simple interest of helping spread the information it contains). Moreover, the mod has no way of knowing whether or not the copyright owner is behind the flag.
Diamond mods should remember, then, that they are an agent of Stack Overflow, but not an agent of copyright holders. But there is something they can do, because there is a separate plagiarism issue: they can make sure the content is attributed properly, by use of a simple edit. This is something anyone with sufficient reputation can do, which is why I say that flagging copyright violations is pointless; most of the time, the correct moderator action is something you could do for yourself: edit to include good attribution. In a nutshell, the terms of use specify that you only post content legally, SE Inc has it's users' promise of that, and this should be good enough until shown otherwise.
This analysis isn't perfect. Some larger sites (ie: YouTube) already have policies and practices that implement preemptive filtering, so obviously it can be done. But I think Stack Overflow is in a different place than is YouTube: YouTube has much more contact with much more litigious copyright holders, and tailored their strategy appropriately.
Moving beyond copyright, any user (including mods) can choose to take a much harder line for plagiarized content, and respond through downvotes, comments, and/or edit deletions — but at this point it's important to remember that, while related, plagiarism and copyright infringement are not the same thing.
The trick is that Stack Overflow blurs the line between infringement and plagiarism because of the authorship issue from my first paragraph: the only way to see plagiarism without corresponding copyright infringement is if the copyright owner were complicit in the plagiarism. On the other hand, it is still easy (though attribution) to have infringement without plagiarism.
Update: tl;dr version:
The Stack Overflow terms of use specify that you only post content legally, SE Inc has it's users' promise of that, and this should be good enough until shown otherwise. Finding the original content elsewhere is not enough to "show otherwise". It proves a copy, but not whether that copy infringes, because we're not in a position to know if the copyright holder might prefer the post be left alone, had made an offline agreement with the poster to allow the post, or any number of other possible exceptions — innocent until proven guilty and all that. There is also a chance that making this choice would open a legal loophole to create some liability for Stack Overflow directly, though while possible I think it's less likely. What we can and should do is respond to the plagiarism issue by at least editing the post to include proper attribution.
