ISPs and Section 512 of the DMCA

What is an ISP?
According to Webopedia, which describes itself as "The only online dictionary and search engine you need for computer and Internet technology definitions," an ISP is:

Short for Internet Service Provider, a company that provides access to the Internet. For a monthly fee, the service provider gives you a software package, username, password and access phone number. Equipped with a modem, you can then log on to the Internet and browse the World Wide Web and USENET, and send and receive e-mail.

It is fair to say that this is what most people think of when they hear ISP: a company that provides Internet service. Most people would also agree that this is not what Google does (yet at least).

Another term is OSP, or Online Service Provider, which can be defined as:

A business that provides its subscribers with a wide variety of data transmitted over telecommunications lines. Online services provide an infrastructure in which subscribers can communicate with one another, either by exchanging e-mail messages or by participating in online conferences (forums). In addition, the service can connect users with an almost unlimited number of third-party information providers.

This sounds more like what Google does, but the catch is that no one "subscribes" to Google. For example, among the largest OSPs are companies like AOL and MSN. Like an OSP, Google connects its users with an almost unlimited number of third-party information providers (like this or your blog), but the openness and free availability of Google's services seems to set it apart from what is traditionally understood as an OSP.

However, Congress has a different definition for what constitutes an ISP under the Digital Millennium Copyright Act.

Why does it matter if Google is an ISP?

Being labeled an ISP has its benefits, namely that section 512 of the Digital millennium Copyright Act provides limitations on liability for copyright infringement by ISPs. By satisfying the various requirements of section 512, an ISP can seek protection from liability for "all monetary relief for direct, vicarious and contributory infringement." 512 also protects ISPs from other things, such as injunctions in certain instances, but the ban on money damages is key for, more often than not, there is little reason to bring a lawsuit if winning won't provide a financial windfall from a defendant with deep pockets.

The Digital millennium Copyright Act, Section 512

The DMCA creates 4 "safe harbors" for parties deemed to be "service providers." These exemptions limit the liability of a party the engages in conduct specified in section 512.

The first safe harbor covers transitory communications. This exemption arises when a service provider is charged with copyright infringement based on the transmission of material through its system.

In order to qualify under the transitory communication exemption, a party must qualify as a "service provider" defined in section 512(k)(1)(A) as:

An entity offering the transmission, routing, or providing of connections for digital online communications, between or among points specified by a user, of material of the user's choosing, without modification to the content of the material as sent or received.

The second safe harbor exemption covers system caching, limiting the liability of a service provider "by reason of the intermediate and temporary storage of material on a system of network controlled or operated by or for the service provider."

The third exemption is for service providers who allow storage of information on systems or networks at the direction of users. Such an example would be a company providing server space for a website or blog.

The final exemption covers information location tools, applying to the "infringement of copyright by reason of the provider referring or linking users to an online location containing infringing material or infringing activity, by using information location tools, including a directory, index, reference, pointer, or hypertext link." This exemption was tailored to specifically address search engines.

In order to qualify for the last three "safe harbor" exemptions, a party must qualify as a "service provider" defined in section 512(k)(1)(B) as:

A provider of online services or network access, or the operator of facilities therefore.

Notice how much broader this definition of service provider is than when dealing with transitory communications. EBay has been declared a service provider under this statute and there is virtually no limit to what could be swept under this definition. Theoretically, any website showing that it provides an online service could qualify as an ISP and avail itself of the liability limitations set forth under section 512.

Impact of DMCA 512
The safe harbor provisions of section 512 exist as additional defenses for search engines, on top of any claim for fair use. That means that even if certain behavior is deemed to be direct infringement, section 512 serves as a limit to the relief sought. However, the ins and outs of how exactly these exemptions apply is still up in the air due to the vagueness of the language in the statute (as noted by the definition of ISP, for example), meaning that much of 512's meaning will have to resolved in the courts. Each of the four safe harbors requires its own post on this blog, detailing the more specific requirements of each and how courts have interpreted them, but that is to come and this post is to serve merely as a framework for further discussion of the issue.

A Filter to Block Copyrighted Images?

Slashdot mentions a new filter called iShield that can automatically recognze porn based on the contents of the image (not the URL or text) and block it from appearing in your browser. If iShield can do this for porn, how long before a similar filter is developed that can censor copyrighted pictures based on content? Would this be good (less lawsuits) or bad (fewer images, especially if such a filter was imposed) for Google and search engines generally?

You Say Tomato, I Say Cease and Desist

Google may run into trouble for linking to other people's sites, arguably because Google is loaded and people hope that it will just pay them off, but what happens if a blogger is sued for linking to another site? Is that even possible?

Anything is possible these days as Neil Gaiman recieved a cease and desist letter from the people who put together the Attack of the Killer Tomatoes movies, asking him not to link to their site without permission. Inconceivable!

The letter he recieved mentions that his links to the Attack of the Killer Tomatoes website are in violation of the Lanham Act (15 USC 1051) and the Copyright Act (17 USC 101).

The letter assumes that Neil operates www.tomatoesaerevil.com, a site "dedicated to the belief that the humble tomato fruit is in fact EVIL," with which he denies any connection, making this letter even more perplexing.

This may be nothing more than lawyers establishing a history of protecting a mark, as well as bad lawyering, but it strikes me as troublesome. Isn't the fair use defense for activity such as Neil's (linking to another site) straightforward enough that such claims can be avoided in the first place? This is a central topic of the paper I'm working on and I haven't found much to support this letter's claim.

Neils's thoughts about the lawyers at issue:

Are they perhaps surrealist lawyers, or cooler than that, Dada lawyers, who have decided to spread artistic confusion and mystery across the web with their "legal letters"?

Another Google Lawsuit Over PageRank

MercuryNews and the AP report that KinderStart, a website aimed at helping parents care for young children, has sued Google for unfairly manipulating its PageRank. Though not a case involving copyright, its worth discussing here because of the implications the case may have over the "automated-ness" of Google's search business and the financial impact of Google's rankings.

KinderStart's lawsuit (find here) will have to overcome the decision in SearchKing v. Google (no longer available online as far as I can tell, what's up with that LawMeme?), a case from three years ago where SearchKing sued over its demotion in Google's search rankings. Google prevailed in that case and central to the decision was the judge's finding that:

While Google's decision to intentionally deviate [sic] from its mathematical algorithm in decreasing SearchKing's PageRank may raise questions about the 'truth' of the PageRank system, there is no conceivable way to prove that the relative significance assigned to a given web site is false. A statement of relative significance, as represented by the PageRank, is inherently subjective in nature. Accordingly, the Court concludes that Google's PageRanks are entitled to First Amendment protection. (page 9)

Thus the court said that PageRanks are opinions, not facts, because it is demonstrably impossible to prove that one website is more relevant than another. Even though Google claims that its PageRank technology is "objective," the court disagreed. But by finding that PageRank is opinion, there is then no room for a plaintiff to claim that it is entitled to any certain search ranking.

Further:

SearchKing consciously accepted the risk of operating a business that is largely dependent on a factor (PageRank) over which it admittedly has no control. The fact that the company with sole control over that factor has unilaterally changed the impact that the factor has on SearchKing's business cannot give rise to a claim for tortious interference with contractual relations. (page 12)

And:

While it could be argued that Google acted maliciously and wrongfully as to SearchKing, the Court concludes that Google's actions were nonetheless privileged [under the First Amendment].

(these quotes were pulled from LawMeme's great summary of the decision. Also read LawMeme's breakdown of the initial suit)

SearchKing was in the business of driving its member sites to the top of Google search result pages by exploiting PageRank features. Such SEO's (Search Engine Optimizers) are paid to elevate pages in search engines. Whether these businesses are legit ("white hat SEO's") or nefarious in their schemes ("black hat SEO's") is not an issue for me right now, but I offer some links to interesting discussions on the issue to satisfy curiosity:

• Google-Opoly: The Game No One but Google Can Play by Dahlia Lithwick (Slate)
• In Contest, Players Vie to Position A Site Atop Google Search by Lee Gomes (WSJ, $ only)
• Google v. Evil by Josh McHugh (Wired)

And here's how to manipulate Google to rankle your friends (or enemies):

• Google Time Bomb by John Hiler (Slate)

There doesn't seem to be much distinguishing KinderStart from SearchKing based on the news reports so far. KindStart may not be manipulating its PageRank to the extent that SearchKing did, but Paul Kedrosky points out that:

Google is the dominant search player, but unless you can demonstrate abusive behavior in how it re-ranks sites, alleging that Google is liable for damages for your lowered position in search results seems like sour grapes from a quasi link-farm.

I thank Paul for his funny and apt description of what this KinderStart page is. Such a page undoubtedly would boost KinderStart's PageRank regardless of the value of the content on its site, and is precisely the kind of behavior Google tries to filter in order to provide its users with relevant search results.

One issue that may distinuish KinderStart's case is the increased web presence of Google, as opposed to three years ago, begging the question of what kind of responsibility Google has in its position of dominance. Jack Schofeld at Guardian Unlimited ponders this point, while describing Google as:

... the police force, sole witness, judge, jury, court of appeal and executioner [when it comes to punishing sites for manipulating the ranking system].

When a site loses 70% of its traffic on Google's say so and is relegated to the "sandbox" (where misbehaving sites go for a timeout) what kind of options does a website have if Google refuses to explain its actions?

Could such unilateral action when the financial stakes are so high be a sign that regulation is on the horizon? Possibly. The Blog Bussiness Summit says that:

Whatever the verdict in such cases, it’s clear that people have come to rely on Google so much that they see it almost as a utility - like hot and cold running water. At some point, that may mean that the search engine industry will become a regulated one.

True, Google has the power to make or break a business, but is that enough for regulation? Water and electricity are regulated not to ensure that businesses can operate, but so that I can shower and cook. Regulation is supposed to protect the public interest and in the case of Google the public interest lies in maintaining the community system of checks and balances (as described on Google's own page explaining how PageRank works) as opposed to shifting the balance towards businesses trying to make a buck by manipulating the system. Of course, once Google's primary business morphs from search to primarily providing directed advertising to web users based on the massive amounts of information it collects (once Google figures out its legal issues with search, I expect Google will help advertisers subsidize free content on the net like they do on TV), regulation may be needed.

Of course, this whole suit could disappear if Google clearly communicated when and why certain sites are demoted. Marketing Pilgrim offers this sensible solution. But in doing so Google may give up too much. Providing such notice would thrust Google into a much more active role in maintaining its search rankings, which could cause problems in defending copyright lawsuits such as in Parker v Google where the court focused on Google's lack of volition.

Not much else of interest up on the web so far, that I've found at least, but the complaint was only filed this week. As a general observation on the case though, I think that Shari Claire Lewis has nailed it in this article on SearchKing when she says:

Businesses must recognize, as reflected in two decisions issued last year by a U.S. district court in Oklahoma, that there are limits to what they can demand without paying for what they want.

Hopefully this KinderStart case will answer the questions of how far Google has to go in defending the demotion of a site and what happens when an innocent site is suspected of PageRank manipulation.

Here's also some other links that may be of interest:

• Businesses Have No Say in Search Engine Rankings, another great analysis of the SearchKing case with a great point that there are limits to what businesses can demand without paying for what they want.
• Google's Giant Sandbox, giving what seems like a good description of the "sandbox" phenomenon.
• How to Play in Google's Sandbox, noting all the ways to get out of the sandbox.
• The Importance of Link Building, learn how to develop good link practices to gain a high search ranking without abusing the system and ending up in the sandbox.
• Bob Massa Interview, read the head of SearchKing discuss SEO's and the SearchKing v Google case.

Clearing the Air: Google Book Search Group

In an effort to clear up misinformation about Google Book Search and the Google Library Project, the folks over at Google started a Google group on the topic. By clicking on the archive at the bottom you can find links to all sorts of informative articles on the subject.

Google Wins Court Battle, Caching is Fair Use

CNet reports that in the case of Parker v Google, the US District Court for the Eastern District of Pennsylvania has dismissed a writer's suit against Google for copyright infringement, agreeing with a Nevada ruling that caching does not constitute infringement.

What seems key is that Google's search technology is completely automated, as the court said:

When an ISP automatically and temporarily stores data without human intervention so that the system can operate and transmit data to its users, the necessary element of volition (willful intent to infringe) is missing.

I've been reading Arribasoft and Perfect 10 in trying to distill the rules for image searches and nowhere have I come across the requirement of a "willful intent to infringe." It would seem unfair for the law to come out one way for pictures and another for text, and even more unfair if courts are picking and choosing which copyright laws they feel like following.

This Court, as the Nevada court in Field v. Google ruled, figured that Google's database falls under the "safe harbor" provision of the DMCA, which protects databases, ISPs, and other online service providers from liability if they don't exert direct control over what content is posted on their sites. Neither court in Arribasoft nor Perfect 10 addressed this issue.

Parker vowed to appeal, stating that the court doesn't understand what caching means and claiming that Google qualifies as a third-party republication.

How Search Engines Corrupt

I think this is classic and certainly funny enough to post.

Lee Gomes wrote an article in the Wall Street Journal a little while ago about how he tried writing for a website and became disgusted with the whole process of creating "original content." He describes the process by which sites plagiarize and manipulate content to get good search placement, which leads to the phenomenon of there being tons of articles on any given subject that aren't very good. By the end of the article he lets loose this description of search engines:

In fact, search engines are more likea TV camera crew let loose in the middle of a crowd of rowdy fans after a game. Seeing the camera, everyone acts boorishly and jostles to get in front. The act of observing something changes it.

Having great search engines is better than not having them, or having bad ones, but Gomes' insights provides a sobering counter to an unflinching devotion to the technological magnificance and social utlity of search technology.

But it's to be expected that no matter what area we are talking about, people will seek to exploit whatever they can to make a buck. Ah, free enterprise!

Headlines, Leads and Copyrights

Wired has a great article hyping what it calls a newsreader development war. Playing up the "man v. the machine" angle, the article describes how different newsreaders operate. From Digg, based on user comments that elevate articles to popularity, to Google News, based on Google's top-secret algorithms, and everything in between, the article points to a burgeoning industry whose aim is to get relevant news into the hands of readers. Unfortunately, there's a lawsuit lurking that could threaten to put the kibosh on these new technologies in AFP v. Google.

Agence France-Presse brought suit against Google (read AFP's complaint and read Google's answer) for copyright infringement and is seeking an injunction based on the fact that Google copies AFP's headlines and displays them on Google News. AFP is a wire service that makes money by licensing its stories, the most important part of which, it claims, are the headlines. Generally, the suit accuses Google of stealing AFP's creative content and distributing it for free, without AFP's permission, and to AFP's financial detriment.

The collective gut reaction is to view the suit as absurd: Headlines can't be copyrighted! William Patry over at Patry Copyright Blog calls the whole mess a big steaming pile of baloney, and offers this legal tidbit:

37 CFR 202.1(a) provides that registration may not be had for:
"(a) Words and short phrases such as names, titles, and slogans; familiar symbols or designs; mere variations of typographic ornamentation, lettering or coloring; mere listing of ingredients or contents."

LawMeme takes a bit more of an activist stance on the issue and calls for blood, claiming that the "case will redefine fair use and the way we link to and use material on the Internet." (Btw, looks like maybe I have to figure out a way to get into BNA to read their stuff).

However, Martin Bishop and Thomas Anderson have a great summary of the case and the issues in the Patent, Trademark & Copyright Journal. While they argue that while headlines are probably not copyrightable, and finding that they are would cause significant problems for the internet, and bloggers especially, they also point out the stark contrast between the two positions. If AFP wins, everyone will have to rephrase headlines on their blogs or get permission (and probably pay $) to use the real ones. If Google wins, AFP may lose its ability to make enough money to create the headlines in the first place. I mean, if I can take an AFP headline and lead paragraph, copy it onto my website (or search engine) and make money on advertising without paying AFP, why would I ever pay AFP?

It seems clear, however, that Google's use of AFP's leads does constitute copyright infringement. So what does this mean? Probably no more than that its ok to lift someone's headline, but if you want to include a lead to entice readers to read the full article, you'll need to write your own. This is what Digg does. Unfortunately, doing so tends to reduce the authoritativeness of the content and encourages plagiarism (read a great story about this by Lee Gomes here).

In another post I will look at the four part fair use test and see how the AFP case could work out under the thinking in Kelly v. Arribasoft and Perfect 10 v. Google. Using someone else's lead may be fair use nevertheless, as could using a headline if the judge here decides they are copyrightable. But I have to get to class now.

Welcome to the Google Copyright Blog

This blog is dedicated to discussing Google and copyright infringement as part of a research project I am writing as a second year law student at New York Law School. Featured items will include analysis on copyright law and how it applies (and maybe should apply) to search and the internet, links to court documents and other postings that are insightful, as well as anything else I find interesting. I hope you enjoy this site and, if you find it useful, please link to it on your site so I can break out of the obscurity of a poor search rating.

If you would like to contact me, my name is Nick Daly and I can be reached at ngdaly at gmail dot com.