Hacker News new | past | comments | ask | show | jobs | submit login

Are you referring to the current state of affairs, or the state of affairs on a hypothetical world without 230? In the world with 230, Facebook is not considered the publisher because they are a "provider" of an "interactive computer service", and the content they are showing is "information provided by another information content provider." Therefore they are not treated as the publisher. That is what s230 does.

In a world without s230, whether they moderate or not, they'd be liable as a publisher for anything their service shows. They would have to dramatically restrict the spread of information to limit risk of litigation, e.g. in the case of people who might sue them for an individual user's slander. Note that this same issue would apply to anyone running newsgroup servers, website hosting, forums, wikis, etc. Republishing UGC would probably become economically challenging.




> the content they are showing is "information provided by another information content provider."

But they aren't. They make decisions about what goes into someone's feed. When they started out and the feed was just a list of items posted by people or groups you followed then it was just 'showing' content.

Using an algorithm to show people content that they will engage with more is not 'showing' something. It crosses the line into publishing. Facebook makes a decision about content


> They make decisions about what goes into someone's feed.

That's irrelevant to the law. The information itself is provided by another information content provider (i.e. whoever posted the status initially). That's who is liable for any damages caused by the content under the law today.

> Using an algorithm to show people content that they will engage with more is not 'showing' something.

All possible mechanisms of showing someone something with a computer involve algorithms. Even if you choose randomly, that is an algorithm. Same goes for sorting by date or alphabetically, since any sort is an algorithm by definition. There's nothing in the law that makes a special case for algorithmic decisions about what to show.

> It crosses the line into publishing.

I'm not sure whether you're expressing a preference here, or a belief about the legal status of Facebook's procedures. In case it is the latter: in fact, it does not cross the line into publishing, legally.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: