2 ms·
§230 is one of the most misunderstood laws in the US, in large part because of intentional efforts to misconstrue it. What it actually says is that websites are
by jcranmer 3y ago
§230 is one of the most misunderstood laws in the US, in large part because of intentional efforts to misconstrue it. What it actually says is that websites are not to be considered the speaker of any user-generated content, and there is explicitly and very intentionally no precondition along the lines of "publisher-platform" or "algorithmic content" or "moderation" or the like.
Now, it's also the case that moderation and content recommendations themselves are speech on the part of the website--crucially so, for that means they also have First Amendment rights to choose those policies without government coercion. There is an interesting question as to whether it is possible for such speech to create liability independent of the underlying user content (for which §230 absolutely immunizes them); indeed SCOTUS heard a case not long ago to answer this question, only to toss it out because the plaintiffs were uninterested in discussing that question.
- SAI_Peregrinus 3y agoThe interesting question is likely (IMO) to be that it can create liability. Selectively picking & ordering the content of others can trivially create new meaning, for example the classic trope of a ransom note formed by clipping words & phrases out of a newspaper & gluing them onto a page. Whether a particular site's editorial actions on user content create liability is likely to be determined on a case-by-case basis. Real site's algorithmic recommendation systems aren't nearly as transformative as newspaper clipping ransom notes, but the sequence and organization of recommended content may well count as a new composite work (much like a music album is a work composed of a sequence of individual tracks) and create liability. I wonder if it would also qualify for individual copyright.