Mike, what they are saying sadly is right. The safe harbors were really intended to protect companies that have no influence over content or it’s presentation, like real hosting companies, connectivity providers, etc.

Instead, what happened is that a whole new class of “service providers” popped up, and suddenly everyone on the internet was a “service provider” instead of a website.

Youtube is a great example. They generate their own HTML, they sort, aggregate, mix, and republish content on new pages, ad advertising, and the like. Yet they want to be treated in the same manner as your ISP who only provides connectivity.

The notification was intended to keep straight hosting companies (those that provide a server and a connection) from being responsible for the content on their servers. Instead, it has been taken to include things like file lockers, which are not hosts in any traditional sense, and who make their money not off of hosting the material, but in charging access for it – and offer the hosting option entirely anonymously.

What was intended, what was understood, and what happened are three totally different things.

So no, they aren’t going back to change DMCA, they are just working to narrow down what a content aggregation service is allowed to do.

So, let’s say I’m running a site that allows users to upload videos. If the law says I’m going to be held liable when one of those videos is infringing, there has to be some way for me to tell whether any given video is infringing or not. Otherwise the law amounts to banning the whole practice of letting users upload anything.

So, given a user-uploaded video, how can the site operator find out whether it is infringing or not?

In the first place, almost everything that qualifies as any sort of expressive work is copyrighted as soon as it’s created. So that doesn’t tell us much.

The next question is who is the copyright owner. If (as a site operator) I make users click something that says they’re the owner or authorized by the owner, is that enough? If not, how can I tell? There is no central registry of copyrighted works. Even if there were, how could a site operator determine whether a video or music file was or was not the same as a work in the list, if the names were different?

Next is the question of whether the upload is or is not authorized by the copyright holder. Again, there is no central registry of such things. It would take all the IT resources in the world to construct and maintain such a vast database of constantly changing information, and presumably some new, mythical technology to match actual files with items in the database.

The companies seeking to destroy the “safe harbor” provisions do not even attempt to explain how any site owner could accept user content with risk of liability, if not for the “safe harbor” provisions. Demanding something impossible as a condition of operating is an order to shut down.

Does techdirt have any way of verifying that this post does not infringe someone’s copyright? If it had to do that, the risk of liability would require it to stop letting users post comments. Every site publishing anything from users would be subject to being bankrupted at any moment by users infriging copyright, because there is no way, even in principle, that it can prevent them doing so.