The legal use of images should be taken care of by the owners of this website. Of course, it would be (very) helpful if the users of the website help in curating the content in that aspect as well. Eliminating weeds at their source is the best. But I don't believe that it is useful when that leads to having moderators/reviewers being restrained in making decisions. For the reviewers the statistical content is the foremost importance. Whether or not an image is 'legal' is welcome to be taken into account, but it is not 'our' responsibility. Don't be too much bothered by it.

The legal use of images should be taken care of by the owners of this website (who, as Scortchi mentions, actually don't do much or anything about it actively). 

Of course, it would be helpful if the users of the website help in curating the content in that aspect as well. 

Eliminating weeds at their source is the best. We can let everything grow and just wait for DMCA take-down notices, but it is a polite thing to act already before that happens.

The least thing that we can do is making sure that there is proper attribution and if some text is unnecessary large then it can be shortened. But I don't believe that it is useful when that leads to having moderators/reviewers being restrained in making decisions. For the reviewers the statistical content (and style readability) is the foremost importance. Whether or not an image is 'legal' is welcome to be taken into account, but it is not 'our' responsibility to judge this at a detailed level. I would say, don't be too much bothered by it.

Aside from legal considerations, there is also some overlap with proper scientific writing, which in my opinion we should promote. Sources for information should be given. And if ideas stem from others then attribution to those others should be given

I have written several comments that do not answer the question, but they have become such large in extent that it might be useful to collect them in a single answer.

I am gonna heavily quote myself here,

which creates a very long answer, and it actually means that I don't want to spend much more time (yet) to create a more concise answer. I might say that I would wish to make this post improved (shorter) but I know by now that I usually don't live up to these promises.

• First, a difficulty with this question is that there might be different contexts and different types/levels of copyright infringement.

  What sort of copyrighted image is it? Is it like a photocopy or is it an image that is available on the internet? Possibly it is sufficient to just add a reference to the creator?

  Also, the solutions may diverge a bit depending on the situation. An edit with an additional image might be large, but possibly you could accept the edit with a re-edit removing the image or changing it to a different one. In other cases removing the image might mute the purpose of the suggested edit.

• Second, there is an issue with the idea whether all images are actually copyright infringement.

  I know two cases where I have been lazy and used figures from others. One example is here where I use a figure from Neyman and Pearson describing their famous lemma. Another example is here where I use a figure from Fisher describing least discriminant analysis. Is such use of images not right? It is possible to quote other people, I thought that such images would fall into this realm as well.

• Third, an extension might be made to the use of code

  Possibly the same is true for the use of code as well. I quote pieces of code from R packages several times and a typical example where I don't even write the reference is here (and the reason is because you get this code very easily).

  A Q&A forum is a bit of a conversation happening on the internet. When I talk (privately) to a person and quote from some article, then I don't need to explicitly reference the source completely. In discussions about a software package, when a specific line of code gets quoted, then I believe it is ok when it doesn't get fully referenced as long as it is clear that the line of code stems from that software package.

Finally, I believe that any legal considerations of an edit should not be a consideration for regular auditors. The foremost question is whether a post is improved or not.

The legal use of images should be taken care of by the owners of this website. Of course, it would be (very) helpful if the users of the website help in curating the content in that aspect as well. Eliminating weeds at their source is the best. But I don't believe that it is useful when that leads to having moderators/reviewers being restrained in making decisions. For the reviewers the statistical content is the foremost importance. Whether or not an image is 'legal' is welcome to be taken into account, but it is not 'our' responsibility. Don't be too much bothered by it.

I have written several comments that do not answer the question, but they have become such large in extent that it might be useful to collect them in a single answer.

I am gonna heavily quote myself here,

which creates a very long answer, and it actually means that I don't want to spend much more time (yet) to create a more concise answer. I might say that I would wish to make this post improved (shorter) but I know by now that I usually don't live up to these promises.

• First, a difficulty with this question is that there might be different contexts and different types/levels of copyright infringement.

  What sort of copyrighted image is it? Is it like a photocopy or is it an image that is available on the internet? Possibly it is sufficient to just add a reference to the creator?

  Also, the solutions may diverge a bit depending on the situation. An edit with an additional image might be large, but possibly you could accept the edit with a re-edit removing the image or changing it to a different one. In other cases removing the image might mute the purpose of the suggested edit.

• Second, there is an issue with the idea whether all images are actually copyright infringement.

  I know two cases where I have been lazy and used figures from others. One example is here where I use a figure from Neyman and Pearson describing their famous lemma. Another example is here where I use a figure from Fisher describing least discriminant analysis. Is such use of images not right? It is possible to quote other people, I thought that such images would fall into this realm as well.

• Third, an extension might be made to the use of code

  Possibly the same is true for the use of code as well. I quote pieces of code from R packages several times and a typical example where I don't even write the reference is here (and the reason is because you get this code very easily).

  A Q&A forum is a bit of a conversation happening on the internet. When I talk (privately) to a person and quote from some article, then I don't need to explicitly reference the source completely. In discussions about a software package, when a specific line of code gets quoted, then I believe it is ok when it doesn't get fully referenced as long as it is clear that the line of code stems from that software package.

Finally, I believe that any detailed and complex legal considerations of an edit should not be a consideration for regular auditors. The foremost question is whether a post is improved or not.

The legal use of images should be taken care of by the owners of this website. Of course, it would be (very) helpful if the users of the website help in curating the content in that aspect as well. Eliminating weeds at their source is the best. But I don't believe that it is useful when that leads to having moderators/reviewers being restrained in making decisions. For the reviewers the statistical content is the foremost importance. Whether or not an image is 'legal' is welcome to be taken into account, but it is not 'our' responsibility. Don't be too much bothered by it.