1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar
Source link: http://archive.mises.org/18568/is-amazons-silk-browser-a-copyright-pirate/

Is Amazon’s Silk Browser a Copyright Pirate?

September 28, 2011 by

Amazon introduced a new tablet today, the Kindle Fire, which uses the “Silk” browser, which is discussed here, and in the video below. One smart thing Silk does to speed up web browsing as seen by the user of the Kindle Fire by “pre-loading” content into Amazon’s “cache” in its own “Amazon computer cloud” (i.e. Amazon’s servers)–and to optimize them for the Kindle Fire (e.g., a 3MB image is scaled down maybe to 50k because that would look the same on the Kindle Fire as a 3MB image, but could be transmitted more quickly). But to do this Amazon’s servers have to store copies of files obtained from other websites, including images (as explicitly stated at 3:07 to 3:26) and other files which, of course, are covered by copyright. At 3:54, it’s explained that if Amazon’s computing cloud sees you looking at the New York Times home page, and it predicts, based on other user statistics, that you are somewhat likely to next click on some NY Times subpage link, then the Amazon servers will go ahead and download that next link, and cache it, in case you do click on it next, so that it can serve it up more quickly.

Now this makes sense technically, but what it really means is Amazon’s servers are making copies of other people’s copyright-protected content: images, files, NYTimes web pages, and serving them up to Kindle Fire users as if the Amazon computer cloud servers are the host of those images. It is a bit like if Amazon ran a site called NYTimes2.com, and had its servers constantly copying content from NYtimes.com and duplicating it on NYTimes2.com, and serving up the content on NYTimes2.com (which was copied from NYTimes.com) to browsers. And who can think that would not summon a lawsuit?

Now, does the idea make technical sense? Yeah. It’s brilliant. Does it infringe copyright? Well, I guess we’ll see! But thank God copyright is there to promote innovation!

Update: In comments on Facebook and elsewhere, several people have asked whether this isn’t just “temporary” or “ephemeral” caching that somehow does not “count” as copyright infringement. I’m of course aware of various provisions of the 1998 Digital Millenium Copyright Act (DMCA) (I taught this in a computer/Internet law class for two years at a local law school in 1998-99 and have written in this general area) ((See my articles “Copyright Infringement and Internet Service Providers,” “Are You Liable and What Can You Do About It” Colloquium for Internet Service Providers, June 11, 1996); “Whither Copyright Law? A Roadmap to Recent Changes in Copyright Law,”The Legal Intelligencer [Philadelphia], Sept. 4, 1997.)) There is a provision of the DMCA that purports to limit ISP liability for the purposes of caching (see 17 U.S.C. § 512(b); and Wikipedia articles on on this section and on web caching). For example, in the 2006 case Field v. Google (discussed here), a federal court ruled for Google when it was accused of copyright infringement because it was caching the plaintiff’s website and making a copy of it available on its search engine.

However, it is not immediately clear to me that the “caching” Amazon plans to do is the same as that done by Google or, in any case, that the “caching” is protected under the statute. Google is providing search results; it is not serving up web pages in place of the original host. And Section 512(b) of the DMCA, which concerns the practice of caching, specifies that “A service provider shall not be liable … for infringement of copyright by reason of the intermediate and temporary storage of material … ” It also provides conditions that must be met for this limitation of copyright liability to apply, including “the material … is transmitted to the subsequent users … without modification to its content….” What is “temporary”? The way I understand the amazon cloud working, it could store sites, pages, and files for months, or even longer. Is that temporary? And if a 50k version of an image is sent to the user instead of a 3MB version, is that a modification of the content? It seems to me that there is enough uncertainty here in the statutory language that some content provider or copyright troll wanting to cause trouble could make an argument against Amazon’s caching practices. Professor Tushnet in the comments below says that Amazon needs to “follow any rules that the originating website has about refreshing content” [which is in Sec. 512(b)(2)(B)] and observes that there is no reason to think Amazon will cache the files “forever.” It’s not clear to me that satisfying the rules for refreshing implies that “temporary” in the statute means “not forever.” If the files for a relatively static site, that changes rarely, are downloaded by the Amazon server, it could be serving up the files for a whole website or web page for months or more. This is compliant with the originating site’s rules on refreshing, perhaps, but is it “temporary”?

Professor Tushnet also says that that “automatic conversion for convenience into a different format or resolution is fine,” i.e., that there is no concern about modifying files, pointing to the case IO Group, Inc. v. Veoh Networks, Inc. Again, she may be right; but that case dealt with Sec. 512(c), not (b), so it’s not clear to me that Amazon’s scaling and optimizing of the stored files satisfies the “without modification” requirement of Sec. 512(b).

Professor Tushnet also says, “You could say that anything is uncertain until specifically litigated. But I don’t think that’s true, and I don’t think people should behave as if it is true; that just lets copyright expand ever further.” I appreciate this, but the problem with legislation as a general matter is that it is almost always inherently vague ((See John Hasnas, The Myth of the Rule of Law.)) and its provisions self-contradictory or contrary to those in other statutes or natural rights. When we have copyright trolls suing left and right, is it possible someone might sue Amazon for copyright infringement despite Amazon having good arguments that they qualify for the DMCA’s caching safe harbor rules? Sure. Is it possible Amazon would lose? I could be wrong, but it seems to me they could.

Now, I suspect Amazon’s lawyers have gone through this and think they have good arguments that their caching is not copyright infringement. But who knows? My point in this post was not that this is definitely copyright infringement–just that there is legal uncertainty, and it “could” be–that’s why I concluded the original post, “Does it infringe copyright? Well, I guess we’ll see!” If it turns out that it is, this just shows another problem with copyright: that it would prevent this kind of great technology. And even if it doesn’t prevent it, until we know for sure, there is legal uncertainty. And, of course, one problem with legislation is that it increases uncertainty (see my “Legislation and Law in a Free Society).

{ 14 comments }

Bryan Okolowicz September 28, 2011 at 2:46 pm

Search engines such as Google have been caching web pages since their inception. Granted, images aren’t cached but I think most people would agree it is not due to any copyright liability. Therefore, doesn’t this practice offer some safety of precedent?

Stephan Kinsella September 28, 2011 at 2:50 pm

Dunno. I guess litigation will tell us. Thank goodness for the regime of legal certainty statutory law affords us!

John P. September 28, 2011 at 3:51 pm

And you had to go and point that out… Shhhh don’t let the copy policy know!

Simon September 28, 2011 at 4:14 pm

Opera Mobile does a similar thing and I havn’t heard of them getting into trouble. Though, it’s good to have some competition — Amazon Silk sounds better.

Stephan Kinsella September 28, 2011 at 4:37 pm

A nym on a blog saying he “hasn’t heard of X getting into trouble for similar stuff” is just as good as a legal opinion, I guess! That settles it!

Rick September 28, 2011 at 8:48 pm

Boy, a little snarky today huh Stephan? No bourbon for breakfast? Or too much?

Erick Malsbury September 29, 2011 at 9:06 am

Copyright law is a sham anyhow.

Ohhh Henry September 29, 2011 at 1:45 pm

I guess litigation will tell us. Thank goodness for the regime of legal certainty statutory law affords us!

This is of course one of the most serious problems with IP – it is such a vague concept that nobody actually knows when an actionable violation has taken place. There are clear violations such as a company mass-copying DVDs and selling them or giving them away, and there are undisputed non-violations such as having your browser keep a copy of web content in its cache. But what about everything else? Such as manually saving a copy of web content in order to read it later, creating a backup copy of a DVD, or making a PVR recording and then letting your friends watch it. Nobody can come up with a simple, clear definition of what is a violation and what is not. This is not very surprising, because the concept of IP is itself vague and undefined.

The only thing about IP that is actually easy and simple to understand is that it is a shakedown of the public by large corporations working in connivance with government.

Stephan Kinsella September 29, 2011 at 1:56 pm

Exactly. This is just right. It’s the reason people who oppose “software patents” are barking up the wrong tree–I’ve written maybe 100 of them myself and am not sure you can define it to exclude it anyway. And yes, IP concepts are inherently vague and ambiguous because they are artificial and not rooted in objective property borders.

I will update the main post above shortly to mention the DMCA provisions on caching, which I’m aware of, but my point was not that this is definitely copyright infringement–just that there is legal uncertainty, and it “could” be; and if it is, this just shows another problem with copyright, that it would prevent this kind of great technology (and even if it doesn’t prevent it, until we know for sure, there is legal uncertainty).

Rebecca Tushnet September 29, 2011 at 5:46 pm

Sorry, but you’re quoting the wrong provision. Transient communications is 512(a), and applies to routing/storage for things like email. Caching, until fairly recently mostly a relic of slower internet speeds, is a different practice, and applying 512(b) to Google’s behavior was a bit of a stretch. This, not so much.

I do agree that copyright in general creates bad uncertainty, but this isn’t one of those instances. The complexity of the provisions here is problematic, and it also raises costs to have to figure out if you’re in the right box, but 512(b) doesn’t require much in the way of judgment calls.

Stephan Kinsella September 29, 2011 at 6:09 pm

Rebecca, I think you’re right–the “transient” word is in 512(a). But it’s still not clear to me that what Amazon is doing for Kindle Fire is covered by 512(b). It specifies that cached copies may not be stored “for a longer period than is reasonably necessary for the transmission, routing, or provision of connections.” It also requires that the content that is stored not be “modified.” Is permanent storing of images, or for months or years, a reasonable period of time, just to transmit and route? I doubt it. And is scaling down images not modifying them? Doesn’t seem obvious to me. Do you have any caselaw construing this? If not–I think my original point stands: this is uncertain.

Rebecca Tushnet September 29, 2011 at 6:19 pm

No, sorry, you’re still quoting 512(a). The coordinate limitation on 512(b) is that the caching entity has to follow any rules that the originating website has about refreshing content. There’s no reason to assume that Amazon violates those rules, or stores the content forever. And indeed, automatic conversion for convenience into a different format or resolution is fine. See IO Group, Inc. v. Veoh Networks, Inc., 586 F. Supp. 2d 1132 (N.D. Cal. 2008).

You could say that anything is uncertain until specifically litigated. But I don’t think that’s true, and I don’t think people should behave as if it is true; that just lets copyright expand ever further.

Stephan Kinsella September 29, 2011 at 7:15 pm

Ah–good point. however, 512(b) does refer to “intermediate and temporary storage of material”. What is “temporary”? The way I understand the amazon cloud working, it could store sites, pages, files, for months, or longer. Is that temporary? Is it “intermediate”? Seems to me there is a non-trivial possibility of litigation.

Stephan Kinsella October 2, 2011 at 7:02 am

These are very helpful comments; I clarified and expanded my updated comments in the main post.

Comments on this entry are closed.

Previous post:

Next post: