April 27, 2007
Utah Legislators Realizing They Screwed Up By Banning Keyword Advertising
By Eric Goldman
Linda Fantin at the Salt Lake Tribune reports on the meeting between Utah legislators and various technology companies (Google, eBay, Microsoft, AOL, Yahoo, 1-800 Contacts and Overstock.com) to discuss the recently enacted Utah Trademark Protection Act banning trademark-triggered keyword advertising.
Based on the SL Trib article, it looks like the Utah legislators are beginning to realize that they got in over their heads (Sen. Eastman's defensive bravado that he "makes no apologies" notwithstanding). For example, the article says that the legislators didn't understand that Utah technology companies 1-800 Contacts and Overstock.com routinely buy other parties' trademarks to trigger ads--even though this is a well-documented fact. Rep. David Clark lamented that "I wish we had had this interaction with industry 60 days ago...We would have all been better off." Great point! The world would be a better place if legislators did their homework first before blasting their legislative guns.
Based on this meeting, it appears the law is in stasis for now. The Utah legislators haven't promised to amend or repeal the law (at least, not yet), but Rep. Clark admitted that "we understand we've got some work to do" and the AOL/1-800 Contacts lobbyist walked away from the meeting thinking litigation wasn't going to be necessary. Meanwhile, according to Sen. Eastman, no efforts will be made to create the registry until further discussions take place.
UPDATE: If you missed it, in an April 11 editorial, the Salt Lake Tribune urged the Utah legislature to unilaterally repeal the law.
April 26, 2007
Quiznos Sued for User-Created Ads--Subway v. Quiznos
By Eric Goldman
Doctor's Associates, Inc. v. QIP Holders, LLC, 2007 WL 1186026 (D. Conn. April 19, 2007)
Hey all you UGC evangelists, listen up! UGC is terrific, but there can be a dark side regarding legal liability. This case provides an example.
Quiznos did a nationwide call for user-generated videos comparing its Prime Rib Cheesesteak with a Subway sandwich. (As a vegetarian, I am kind of hoping Quiznos loses as punishment for its bad carnivore karma). To stimulate submissions, Quiznos posted 3 user-generated video examples. Subway complained that these examples contained false and misleading statements; ultimately, Subway sued Quiznos for false advertising under the Lanham Act.
Quiznos moves to dismiss the false advertising claim per 47 USC 230, saying that the statements in the ads were made by Quiznos' users, not Quiznos itself. In a highly technical ruling, the court says no because it views 230 as an affirmative defense that can't support a motion to dismiss. This is 100% wrong and completely inconsistent with the vast bulk of the cases that have, in fact, used 230 to dismiss lawsuits based solely on the complaint (see, e.g., the flagship Zeran case, which was a judgment on the pleadings). Indeed, in both the Doe v. Bates and the UCS v. Lycos cases, the courts granted 12(b)(6) motions to dismiss specifically to shut down fishing expeditions in discovery.
Even though Quiznos didn't knock out the false advertising claim now, it still can raise the defense later--though it's not clear it will qualify for the defense.
In a footnote, the court says that Subway can get discovery on whether Quiznos altered or was creatively involved in the users' videos. In general, 230 will still apply even if Quiznos altered and provided creative input into the videos, but 230 doesn't apply if Quiznos in fact authored the statements in question. See, e.g., here and here. We'll need more facts to resolve this.
Further, 230(e)(2) says that the 230 immunization isn't available for "intellectual property" claims. Is "false advertising" an IP claim? Ordinarily I would say no. See the 2004 district court ruling in Perfect 10 v. CCBill, which expressly held that a state false advertising claim was preempted by 230. So Quiznos should get 230 protection for user videos. However, Subway is bringing suit under the false advertising provision in the federal Lanham Act--the same act that governs trademark law generally. Given the language's location in the Lanham Act, sitting shoulder-to-shoulder with trademark infringement restrictions, I could see some courts viewing it as an IP law outside the scope of 230. We'll see.
On a personal note, if you want to check out the fun I had at Epinions with user-generated commercials waaay before they were hip, see my story and reviews/screenshots of the commercials (iMac, Tent, Minivan, Breast Pump, Alta, George Foreman Grill and Beer).
April 25, 2007
Academic Debate over Trademark Use in Commerce
By Eric Goldman
As regular blog readers know, a hot area in online trademark law is the "trademark use in commerce" element of the plaintiff's prima facie case. This element has been dispositive in several noteworthy defense wins, including the 1-800 Contacts v. WhenU and Rescuecom v. Google cases. Yet, some experts, such as Prof. McCarthy, believe that the prima facie case doesn't require the plaintiff to show that the defendant made a trademark use in commerce.
This debate has led to 2 companion articles by 4 top trademark scholars. First, Graeme Dinwoodie and Mark Janis wrote Confusion Over Use: Contextualism in Trademark Law. In the article, Dinwoodie/Janis agree with McCarthy that trademark use in commerce isn't a separate element of the plaintiff's prima facie case because, otherwise, the trademark fair use statutory provision makes no sense. Therefore, Dinwoodie/Janis favor a contextual analysis of infringement which effectively omits the trademark use in commerce element and instead focuses on the likelihood of confusion factor.
Mark Lemley and Stacey Dogan responded with Grounding Trademark Law Through Trademark Use. This article argues that the trademark use in commerce element historically helped distinguish between direct and contributory infringers (i.e., if the defendant hadn't made a TM use in commerce, its only liability would be contributory).
As I've said before, I think the statute is unresolvably irresolute, so statutory arguments aren't likely to help us reach a consensus any time soon. From a normative standpoint, I think Lemley and Dogan have the better argument in this debate, but for different reasons than the ones they articulate. I think the use in commerce requirement acts to keep "commercial referential uses" outside the boundaries of trademark law. I'll explain this in more detail soon.
The Dinwoodie/Janis abstract:
This paper tackles an intellectual property theory that many scholars regard as fundamental to future policy debates over the scope of trademark protection: the trademark use theory. We argue that trademark use theory is flawed and should be rejected. The adoption of trademark use theory has immediate practical implications for disputes about the use of trademarks in online advertising, merchandising, and product design, and has long-term consequences for other trademark generally. We critique the theory both descriptively and prescriptively. We argue that trademark use theory over-extends the search costs rationale for the trademark system, and that it unhelpfully elevates formalism over contextual analysis in trademark law rulemaking. The theory seeks determinate trademark rules in order to encourage a climate of certainty for innovators, but the concepts on which it is founded are likely to degenerate. We show that trademark use theorists ignores the multivalence of trademark law, and that adopting trademark use doctrines would result in less transparent trademark decisionmaking. Instead, we propose that trademark law retain its traditional preference for contextual analysis. We show in particular how a contextual analysis would offer an approach to trademark disputes involving online advertising that better captures the potential of trademark law to police new information markets. Our analysis contemplates individualized assessments according to common law standards, but opens up policy space for the development of limited statutory safe harbors for intermediaries such as search engines.
The Lemley/Dogan abstract:
The debate over trademark use has become a hot-button issue in intellectual property (IP) law. In Confusion over Use: Contextualism in Trademark Law, Graeme Dinwoodie and Mark Janis characterize it as a dispute over whether to limit trademark holder rights in a new and unanticipated way. Yet there is another – in our view more historically accurate - way to frame the trademark use debate: the question is whether courts should, absent specific statutory authorization, allow trademark holders to assert a new and unprecedented form of trademark infringement claim. The pop-up and keyword cases involve attempts to impose third-party liability under the guise of direct infringement suits. Dinwoodie and Janis's thorough account notwithstanding, it remains the fact that, before the recent spate of Internet-related cases, no court had ever recognized a trademark claim of the sort that trademark holders are now asserting. Trademark infringement suits have always involved allegations of infringement by parties who use marks in connection with the promotion of their own goods and services. The question raised by the trademark use cases, as we view it, is whether courts should countenance a radical departure from that traditional model without specific instruction from Congress. We think they should not.
In this paper, we explain the origins of trademark use doctrine in traditional limits on the scope of the trademark right and in the distinction between direct and contributory infringement. We also explain why we cannot simply rely on the likelihood of consumer confusion test to solve the problems the trademark use doctrine addresses, and we examine the difficult problem of defining the scope of the trademark use doctrine.
New(ish) Report on 512 Takedown Notices
INTELLECTUAL PROPERTY AND FREE SPEECH IN THE ONLINE WORLD: How Educational Institutions and Other Online Service Providers Are Coping with Cease and Desist Letters and Takedown Notices by Laura Quilter and Marjorie Heins
You may recall that Laura Quilter co-wrote an excellent 2005 study on 512 takedown notices. This time, she's cowritten an ethnographic-style study of how takedown notice recipients handle the notices. The report does an excellent job explaining how copyright owners are, in essence, spamming service providers with low-merit takedown notices, which in turn imposes significant investigation and remediation costs on service providers to assess the merit of the notice and act on it. Naturally, for many service providers, the lowest-cost path is not to do individualized investigations but to treat a takedown notice as a presumptive veto of the content.
We all know that the 512 takedown notices was designed to allocate remediation costs between content owners and service providers, but the spammy nature of the content owner's takedown notices has reduced their costs by shifting more of the costs to the service provider. This situation was supposed to be remedied by 512(f), but after the Ninth Circuit Rossi case, 512(f) doesn't have much punch, so it has done little to curb the takedown spamming by content owners. This situation may need further legislative correction to revitalize 512(f) or to otherwise create some disincentive to reduce takedown spam.
The report's five main findings:
1) "large educational institutions and other service providers are swamped by notices about peer-to-peer filesharing (“P2P”) and “abuse” complaints relating to spam, viruses, phishing, and network security," so all notices are processed on a mass-volume and lowest-common-denominator basis.
2) "the flood of P2P notices places significant burdens on service providers"
3) "when in doubt...institutions tend to take a much more restrictive stance than is warranted by the law."
4) "this restrictive stance is driven in part by the confusing nature of the law and the lack of solid information and model policies for service providers."
5) "the situation for users is grim."
HT: EFF Deep Links.
Google Sued for Defamatory Search Result--RSA Enterprises v. Bad Business Bureau
By Eric Goldman
RSA Enterprises v. Bad Business Bureau, No. 2:07-cv-01882-HAA-ES (D.N.J. complaint filed April 23, 2007)
Bad Business Bureau, a/k/a the Rip-off Report, is a well-known repeat Internet defendant (see this article from the Phoenix New Times explaining why). They are a rare two-time loser of the 47 USC 230 defense (see here and here) because plaintiffs have alleged that the Rip-off Report, and its founder Ed Magedson, actually write or embellish griping/bashing content themselves rather than publishing griping/bashing content from users. If the content in question came from users, the Rip-off Report is clearly protected by 47 USC 230 and lawsuits against them for such content should be clear losers. On the other hand, if Rip-off Report writes the content itself, the issue gets more complicated.
In addition to suing the Rip-off Report, this particular plaintiff named Google as an additional defendant for disseminating the allegedly defamatory content as part of its search results. Bad move. Google is unquestionably covered by 47 USC 230 for this content, and there is no way the plaintiff can get around the statutory immunization. So the only question in my mind is what sanctions will attach to the plaintiff's poor choice. Will Google get Rule 11 sanctions? Does New Jersey provide anti-SLAPP relief? (See the New Jersey page on the California Anti-SLAPP Project, and note that Google got anti-SLAPP relief in a similar CA case involving Mark Maughan). With any luck, this case will provide an excellent precedent to establish why plaintiffs should leave search engines alone for their search results.
April 24, 2007
Referencing TM in Online Ad Copy May Be TM Use in Commerce--Hamzik v. Zale
By Eric Goldman
Hamzik v. Zale Corp./Delaware, 2007 WL 1174863 (NDNY April 19, 2007)
Another lawsuit against an advertiser for buying keyword advertising. In this case, the plaintiff claims a trademark in the term "Dating Ring" and says that Zales.com buys the keyword "Dating Ring." Zales.com responds that it simply broad-matched the keyword "ring." The court sidesteps the broad-matching issue but agrees with the defendant that simply buying a keyword isn't a trademark use in commerce, joining the Merck court (also from a Second Circuit-based district court) in holding that such "internal uses" by advertisers don't qualify as a TM use in commerce. However, the court then distinguishes the Rescuecom and Merck precedent because Zales.com's ads displayed the term "dating rings" in the ad copy. The court says that this may constitute a trademark use in commerce, so the case survives defendant's motion to dismiss.
1) This court is right to consider the actual ad copy rather than trying to manufacture a blanket rule governing all different implementations of keyword advertising. However, two caveats: (1) not all trademark references in the ad copy should qualify as a trademark use in commerce; for example, IMO, a comparative reference isn't a TM use in commerce even if it's made in the ad copy--this is a "commercial referential use" that I will blog on soon, and (2) by requiring the judge/jury to look at the actual ad copy, this can make the lawsuit more expensive for both parties due to the costs of factual investigations, proof and adjudication.
2) This ruling implicitly supports the wisdom of Google's policy allowing trademark owners to block TM references in the ad copy (the GEICO case also implicitly validated the policy).
3) This is yet another case (see also the Picture It Sold case) where the court doesn't seem to understand broad-matching. It wouldn't have changed the results of this case, but at some point courts will need to get it!
April 21, 2007
Best and Worst Internet Statutes
By Eric Goldman
At InformIT, I published an article ranking the best and worst Internet laws. The abstract: "Over the past dozen years, the lure of regulating the Internet has proven irresistible to legislators. In the spirit of good fun, Eric Goldman offers an opinionated list of personal votes for the best and worst Internet statutes in the United States."
This list was fun to write, and I hope it stirs up some discussion about how we assess and rank legislative efforts to regulate the Internet. Before you check out the article, see if you can guess what I think are the best and worst laws, and let me know if my list surprised you (or go ahead and leave a comment there). If it makes a difference to your guesses, the Utah anti-keyword ad law came out while the article was in InformIT's editing cycle, so I was only able to include a brief reference to it--I would probably rank it the second-worst law, substituting it for the current #2, although the law is so bad that it does put pressure on the #1 worst law's ranking.
April 19, 2007
Google Gets Mixed Bag in Latest Ruling in American Blinds Case
By Eric Goldman
Google, Inc. v. American Blinds & Wallpaper Factory, Inc., C 03-5340 JF (N.D. Cal. April 18, 2007)
The lawsuit between Google and American Blinds has been running for 3 1/2 years now, a very long time in the Internet world. Yesterday, Judge Fogel (who also handled the KinderStart and Person lawsuits against Google) issued a technical omnibus ruling that gave a little to each side--exactly the kind of ruling designed to facilitate settlement. We'll see if the parties take the hint.
The main rulings:
1) Google asked the judge to reconsider if selling keywords constitutes a trademark use in commerce. In March 2005, the judge said it did, but Google pointed to the intervening Merck and Rescuecom decisions saying that it wasn't. The judge did the best he could to divine the Ninth Circuit's intent from its garbled Brookfield and Playboy v. Netscape precedents. Based on those, he ultimately concludes that the Ninth Circuit implicitly had concluded in those cases that keyword usage and keyword triggering, respectively, constituted a trademark use in commerce. Thus, he reaffirmed his holding from March 2005. While this is hardly surprising, it does raise a more explicit specter of a circuit split between the Second and Ninth Circuits on what constitutes a trademark use in commerce. We should learn more from the Rescuecom appeal, though the Second Circuit is hardly a rocket docket.
2) The court holds that the terms "American Blinds" and "American Blind" are descriptive, and American Blinds hasn't shown secondary meaning, so American Blinds has no trademark rights in the terms. Thus, not only does the court dismiss American Blinds' claims based on Google using those terms as ad triggers, but American Blinds is going to have an uphill battle enforcing those marks in the future. In this respect, the judge's ruling is a significant loss for American Blinds. Further, I haven't seen the data, but I suspect a big number of the complained-about ads were triggered by these two terms as opposed to American Blinds' other TMs ("American Blind Factory," "Decoratetoday," and "American Blind & Wallpaper Factory"). If so, as a practical matter, this ruling might substantially depress American Blinds' hope of getting big damages or an injunction covering lots of competitive behavior. Therefore, by narrowing the amount at issue, this ruling might further encourage the parties to settle.
Despite this ruling, under the new Utah anti-keyword advertising law, American Blinds likely would be able to register the non-trademark "American Blinds" and block competitive keyword advertising.
3) Google argued that it made no sense to apply the Sleekcraft multi-factor test to its behavior. This argument is absolutely correct as a matter of logic, but the judge is bound by the silly Playboy v. Netscape precedent where the Ninth Circuit did a (very messy) Sleekcraft analysis on very analogous facts. Here, the court addresses a few of the Sleekcraft factors and says they generally apply against Google in the context of rejecting Google's request for summary judgment. Therefore, if the parties don't settle, there should be a trial on consumer confusion on the remaining three American Blinds trademarks.
4) The court says that American Blinds' trademarks aren't famous or distinctive, so the court dismisses the dilution claims.
As you can see, each party got a little from the judge here, as he significantly narrowed the points of dispute. Normally, a mixed ruling like this can help the parties reach a settlement. We'll see what the parties do.
Meanwhile, this case remains a flagship example of a lawsuit that, from the trademark owner's standpoint, absolutely makes no financial sense. American Blinds' legal bills for this lawsuit surely must be in the hundreds of thousands of dollars, and I simply can't imagine that American Blinds has lost that much profit from customers "diverted" by the keyword advertising. Further, this lawsuit has also established that American Blinds has no protectable trademark in the term "American Blinds," so American Blinds has also "lost" that asset. All the more reason for them to cut their losses now.
April 18, 2007
Judge Kozinski Talks About Cyberlaw
By Eric Goldman
Last October, Judge Alex Kozinski of the Ninth Circuit chatted with me in my Cyberspace Law course for 75 minutes. If you listen to the recording, you'll hear Judge Kozinski's humorous thoughts on receiving gifts when he speaks (he auctions them on eBay), selling on eBay (he's very proud of his feedback rating), blogging (he hates bloggers--told me that right to my face at breakfast after I told him I was a blogger) and being a judicial male "superhottie," as well as a less spirited discussion about the law of virtual worlds and the Sex.com case (it's hard to pin down sitting judges on substantive legal doctrines). Hope you enjoy the interview.
April 12, 2007
Keyword Law Talk
By Eric Goldman
Tomorrow I'm giving a talk in Dallas entitled "Keyword Law." In light of the new Utah law, this turns out to be an interesting time to speak on the topic!
April 09, 2007
Keyword Advertising as Corporate Identity Theft—Sen. Eastman Defends New Utah Law Banning Keyword Advertising
By Eric Goldman
Last month, Utah quietly passed SB 236 to outlaw keyword advertising. Last week, SB 236’s sponsor, Sen. Dan Eastman, blogged a defense of the law. I really respect Sen. Eastman for engaging his critics in the blogosphere, but his response also illustrates how Utah passed such a rotten and anti-consumer law.
Sen. Eastman’s blog post is entitled “Identity Theft: The Next Generation,” and he explains that keyword advertising is a “creative new kind of identity theft.” Apparently, then, banning competitive keyword advertising should prevent a type of corporate identity theft. While invoking the Big Scary Threat of “identity theft” is a clever rhetorical move, it’s also analytically indefensible. Identity theft occurs when someone makes a false representation, but this law bans competitive keyword advertising that is completely truthful and does not confuse anyone.
Along the way, Sen. Eastman picks some colorful verbs to describe competitive keyword advertising, analogizing competitive keyword advertising to “carjacking” someone’s trademark (should we call this “markjacking”?) and suggesting that searchers presented with a comparative ad are being “shanghaied by a pirate.” These chosen verbs imply that somehow searchers are being forcibly deprived of their rights, as if a thief is holding a gun to the searcher’s head, saying “GIVE ME YOUR CLICK,” or that a careless searcher will wake up one morning with a bad headache trapped on a square-rigged ship as an indentured servant. But, there is no compulsion taking place with keyword advertising; searchers aren’t being forced to do something they don’t want to do. Search engines present keyword advertising, and searchers click on it when they think it’s relevant. Search engines increase consumer choice, not limit it, so in my opinion the only people being shanghaied here are Utah citizens harmed by this law and its bogus justifications.
Sen. Eastman also declares: “I make no apologies. Utah is a highly tech-savvy, super business-friendly state. We have more computers per capita than anywhere else in the nation.”
I can’t argue with his business-friendly characterization, as this law amply confirms that Utah happily sells out consumers to help businesses seeking to limit fair and legitimate competition. However, his reference to the number of computers in Utah is more puzzling. Utah residents may own a lot of computers, but the Utah legislature appears to be trying to render those computers useless by passing a dizzying array of regressive, anti-consumer, anti-Internet laws.
Finally, Sen. Eastman’s intransigence (“I make no apologies”) is understandable but unfortunate. This law will fail in the courts, and it would be a true public service to declare a “mea culpa” than to waste a lot of Utah taxpayers’ money in a futile defense of the law.
In addition to Sen. Eastman's post, Matthew Prince, CEO of Unspam, also blogged his own defense of the law. I'm still trying to parse Unspam's interest in this law--are they hoping to become the electronic registration mark's database vendor?
In any case, I thoroughly dissected Prince's proffered rationales (such as the bogus claim that Mazda "diverts" searchers for Pontiac) in this article, so I'll spare you the rehash here. However, I will address one legal point: it's misleading to suggest that the "Trademark Protection Act merely extends the same rights already enjoyed by mark holders throughout the rest of the world to Utah." Just like US courts, foreign courts are struggling to determine the proper application of their trademark laws to competitive keyword advertising, so there are numerous foreign rulings validating the legitimacy of competitive keyword advertising (see, e.g., Israel's endorsement of the practice in the Matim Li v. Crazy Line case).
UPDATE: It appears that Matthew Prince knows more about the dormant commerce clause than Utah's General Counsel. If this law isn't generating profits for Unspam, it's very generous of Unspam to give Matthew so much time to write lengthy legal memoranda.
March 2007 Quick Links Part 2
By Eric Goldman
Yesterday I posted the Google edition of my list of interesting items from March. Today I post the remainder of items that caught my eye last month.
* Bosley Medical Institute v. Kremer, 2007 WL 935708 (S.D. Cal. Mar. 22, 2007). On remand from the Ninth Circuit, the district court denies Kremer's motions to dismiss/for SJ. Michael Atkins recaps the ruling and case's history.
* Milbank Tweed Hadley & McCloy LLP v. Milbank Holding Corp. d/b/a Milbank Real Estate Services, No. CV 06-187-RGK (JTLx), (C.D. Cal. Feb. 23, 2007). After passage of the Trademark Dilution Revision Act, the court rejects the existence of "niche fame" as support for a dilution action. I’m a little surprised that this plaintiff would bring this losing argument.
* ICANN votes down a .XXX TLD. Again.
* NYT on the increasing challenges of creating a unique global brand in very crowded namespaces.
* Trademarked Sentences: A tool that helps you generate poetry by mixing trademarked slogans.
* BidZirk v. Smith, No. 06-1487 (4th Cir. March 6, 2007). The Fourth Circuit, in a non-substantive opinion, denied a company's request for an injunction against a griping blogger's use of its trademarks. My initial write-up of the case. With this loss, the plaintiff's ill-advised decision to appeal the case is now even more clearly a complete waste of the plaintiff's money and our judicial resources.
* Chapman v. Merchandise Mart Properties, 2007 WL 922258 (D. Vt. Mar. 23, 2007). Woman tries to get TRO against physical-space trade show based on trademark interests in the term "GreenStyle," which is her blog’s title. The court rejects the request, but interestingly doesn't seem fazed by the argument that she may have a trademark interest generated from her blog name. Blog names can be trademarkable with sufficient use in commerce, a factor the court ignored completely.
* Sifry: "70 million weblogs. About 120,000 new weblogs each day, or...1.4 new blogs every second."
* A nice retrospective on the history of blogging.
* Wikipedia is requiring some credentialing after getting burned by a pseudonymous contributor who falsely claimed he was a professor.
* Ed Felten has some terrific observations about building distributed reputation systems like Digg (and, for that matter, Epinions). Ed is 100% correct that reputation systems need substantial stabilization; they don't just work deus ex machina.
* Dorr v. Yahoo, No 3:07-cv-01428-MJJ (N.D. Cal. complaint filed March 7, 2007). Yahoo offered a premium subscription service allowing users to send email without Yahoo's ads attached. Then, allegedly, they changed the service's terms, and some of the paying customers were unilaterally bumped to a tier where Yahoo's ads were again attached to their email. Steve Bryant has more. In general, if people pay to eliminate ads, during that period of time, Yahoo should not be able to unilaterally amend the terms so that the user is paying but still getting ads.
* Ken Adams blogs on Affinity Internet, Inc. v. Consolidated Credit Counseling Services, Inc., 920 So. 2d 1286 (Fla. Dist. Ct. App. 2006), where the court held that a contract clause saying "This contract is subject to all of SkyNetWEB's terms, conditions, user and acceptable use policies located at http://www.skynetweb.com/company/legal/legal.php" was insufficient to incorporate an arbitration clause contained in the referenced document. Ken's suggested fix: "The SkyNetWEB user agreement located at http://www.skynetweb.com/company/legal/legal.php constitutes part of this agreement."
* The National Do Not Call Registry: Annual Report to Congress for FY 2006 Pursuant to the Do Not Call Implementation Act On Implementation of the National Do Not Call Registry (April 2007): "The Commission believes that the fundamental goal of the National Do Not Call Registry — to provide consumers with a simple, free, and effective means to limit unwanted telemarketing calls — has been realized." My curmudgeonly take on why the do-not-call registry isn’t great policy.
* Implementing the Children's Online Privacy Protection Act: A Federal Trade Commission Report to Congress (February 2007). The FTC remains pretty pleased with itself about COPPA, but it's worried about social networking sites and the continuing lack of age verification technology. I'm not as impressed with COPPA as the FTC is; see here and here. In any case, if you're doing COPPA research, this report helpfully recounts the 12 COPPA enforcement actions to date.
* Terrific post by the EFF’s Seth Schoen about a misguided report on P2P file sharing by the USPTO and the issues with empowering users to control their computers. A must-read.
* ACLU v. Gonzales, No. 98-559 (E.D. Pa. March 22, 2007). On remand from the Supreme Court, the court once again holds that the 1998 Child Online Protection Act is unconstitutional.
* CRS Report for Congress: An Overview of Recent U.S. Supreme Court Jurisprudence in Patent Law, March 16, 2007, discussing the last 8 Supreme Court patent cases.
* We've all heard about the magic of network effects. But as this Mercury News article explains, when an Internet start-up company's network takes root principally overseas, it can leave the company with a large audience of unmonetizable users.
* Jacob Loshin, Property in the Horizon: The Theory and Practice of Sign and Billboard Regulation, 30 Environs 101 (2006). A thoughtful discussion of the history of billboard regulation and some regulatory considerations.
* Coca-Cola's launch campaign for "Coke Zero" is premised on the idea that the executives of Coca-Cola want to sue the executives of Coke Zero (i.e., other executives within the same company) for "taste infringement" because the taste is so similar. Personally, I find commercials about faux lawsuits HILARIOUS. Ha ha ha. Except...if there isn't currently a cause of action for "taste infringement," with the expansion of IP rights, it may only be a matter of time... This turns the joke about how hard it would be to establish taste infringement on its head. Ironically, the commercial features Coke's actual lawyers. Yet more on this sorry story.
April 08, 2007
March 2007 Quick Links, Google Edition
By Eric Goldman
My monthly recap of interesting news items grew too big this month, so I'm breaking it into two postings. With so much attention directed towards Google, this first edition of March's quick links will focus just on Google-related items.
* Google's litigation counsel, Michael Kwun, gave a talk called "Keyword-Targeted Ads: The New Trademark War" at SMU in late February. His slides. I trust it is clear that these slides aren't intended to be a neutral description of the law!
* Google won a case in Belgium over Google Suggest, a feature that automatically suggests search terms as searchers type their query into the search box. In this case, the plainitff was a software vendor upset that the suggested search terms prompted users to find cracked versions of its software. See InfoWorld, Techdirt and Search Engine Land writeups.
* Matim Li v. Crazy Line (Tel Aviv District Court July 31, 2006). I normally don't blog on foreign trademark keyword cases but I finally got my hands on an English translation for this case. It's a nice win for Google and it has some interesting discussion.
* At Santa Clara University School of Law in January, we had a panel on the Perfect 10 v. Google case featuring Prof. Justin Hughes of Cardozo, Andrew Bridges of Winston & Strawn, Russ Frackman of Mitchell Silberberg, Fred von Lohmann of EFF and Prof. Tyler Ochoa of SCU. This turned into a major copyright geekfest! Watch the video (free registration required).
* For a while, Ask.com searchers using the search term "Google" got the following result: ""Don't be a droid -- use different sources of information" next to a drawing of a man on puppet strings and a link to Ask.com's anti-Google Web site. Nice…
* Eric Schmidt of Google: "I think the most important characteristic of an entrepeneur is that they’re going to do it whether you give them permission or not." A dangerous quote when you're in the middle of numerous IP lawsuits! HT Search Engine Land.
* BusinessWeek: Is Google Too Powerful?
Oracle v. SAP Lawsuit Comments
By Eric Goldman
Oracle Corporation v. SAP AG, 3:07-cv-01658-EMC (N.D. Cal. complaint filed March 22, 2007)
I realize I'm a couple weeks late to this story, but it's too important/interesting a case not to address.
TomorrowNow (TN) is a company started by former Oracle employees. They offer maintenance services for Oracle software competitive with Oracle's standard maintenance program, but at much-reduced prices: Oracle charges 22%/yr while TN charges half that (11%/yr).
But how is TN able to undercut Oracle's pricing so drastically? One possibility is that Oracle charges supra-market rates due to the lock-in effects of tying maintenance services to software licenses. On that front, I'll note that back in the 1990s, my software vendor clients typically charged 15%/yr for maintenance--a substantially lower number than Oracle's breath-taking 22% figure. So perhaps TN is able to charge 11% as a modest start-up discount off the industry-standard 15%, and Oracle's been getting away with a great deal for a long time.
An alternative story, told by Oracle in its complaint, is that TN could undercut Oracle only by stealing. TN has a very thin development team compared to the Oracle behemoth, so Oracle might incur all of the development expenses necessary to provide maintenance services, and TN might just take those assets for free to engage in competitive free-riding. Specifically, Oracle alleges that TN gets switching Oracle customers to give TN their passwords to Oracle's website/database for its maintenance customers and then send robots to download everything (manuals, patches, etc.) it can find, which then allows it to provide services comparable to Oracle’s.
Perhaps TN, even if had engaged in such a scheme, would have been a nettlesome gnat as a standalone company, but it got scooped up by German software giant SAP, one of Oracle's main rivals. At this point, TN becomes problematic to Oracle in a variety of ways. TN is poaching some maintenance revenue outright, while it is putting price pressure on the maintenance business that Oracle retains. Further, Oracle customers who switch to TN have an easier path to migrate their overall software needs to archrival SAP.
Oracle has struck back in court with a tightly drafted complaint. Oracle claims that the scheme of getting former Oracle customer passwords and downloading lots of content from Oracle's maintenance database violates (among other things) the CFAA and Cal. Penal Code Sec. 502 and constitutes trespass to chattels and interference with prospective economic advantage. This is a well-pleaded complaint, in the sense that there are no obvious deficiencies with Oracle's pleadings. I don't love everything about Oracle's practices. For example, it makes no sense that Oracle made it possible for customers to root around the entire database for stuff, even if it didn't relate to the customer's software. Also, I would definitely have drafted and implemented the contracts differently than Oracle did. But these are quibbles; Oracle's contracts and practices are serviceable for this lawsuit's purposes.
Having said that, there are two obvious omissions from the alleged claims. First, Oracle didn't allege copyright infringement yet because it needed to get its copyright registration applications on file, so it expects to file an amended complaint. Second, Oracle didn't allege that the claimed misuse of switching customers constituted a 1201 circumvention. I'm not 100% sure why. It could be that this claim will be added along with the other copyright infringement claims, or it could be that Oracle is sufficiently deterred by the handful of cases holding that mere misuse of a legitimately issued password isn't a circumvention.
Also, it's noteworthy that Oracle didn't sue its switching customers for allegedly providing their passwords to TN, although it seems like at minimum Oracle would have breach of contract claims against them. I assume Oracle isn't suing customers because that's never good for business. Indeed, part of the lawsuit is about wooing customers; there is some hilarious and gratuitous marketing language in the complaint designed to impress Oracle customers and to rattle the confidence of customers thinking of switching to SAP.
Putting aside what's not in the complaint, if Oracle's complaint accurately states the facts, SAP could be in deep legal trouble. Of course, it's fairly typical for the plaintiff to draft a great complaint and the defendant then tells a very different story. As just one example, Oracle ties the downloads to TN via IP addresses; but IP addresses are spoofable, so it's theoretically possible that someone spoofed TN. So we have to wait until we hear both sides before we can make any rigorous assessments of merit.
Even so, I'm a little unnerved by the software industry analysts who have claimed this lawsuit is no big deal. Perhaps in the grand scheme of things, this lawsuit won't have a great deal of effect on the competitive position of SAP and Oracle. Sure, the lawsuit casts some doubts in the minds of customers who are thinking of leaving Oracle for lower-cost options that SAP/TN will be a long-term trustworthy vendor, but such doubt-sowing initiatives are fairly common the bare-knuckle competition for enterprise database software. Plus, if SAP just cuts off TN altogether, presumably the overall effect on SAP and Oracle revenues will be comparatively modest.
But this lawsuit could be a Big Deal because the facts alleged by Oracle might support criminal prosecutions for CFAA, CA Penal Code 502, criminal copyright infringement and other crimes. It's not clear if the criminal prosecutors are going to get involved in this case or if Oracle even wants them to do so, but I suspect a number of SAP employees have procured their own personal attorneys. To the extent TN was a rogue operation operating without oversight or permission from SAP corporate, then again the financial impact may be small, even if the affected individuals might suffer severe consequences. But if TN wasn't a rogue operation, any criminal prosecutions could have major ripple effects throughout the entire SAP organization.
I think the Cadence v. Avant lawsuits are illustrative, especially given the many parallels. In that case, a bunch of former Cadence employees started up a competitive company, Avant. However, to get a jumpstart on the competition, the employees walked out the door with Cadence source code. Perhaps aided by this unfair head start, Avant had a very successful marketplace run, growing into a major public company with hundreds of millions of dollars of revenue. But after the civil and criminal prosecutions, Cadence got damage awards of hundreds of millions of dollars, multiple Avant employees went to jail, and Avant was effectively knocked out of the marketplace.
I need to reiterate that we don't know yet if Oracle's alleged facts are true, or if anyone committed a crime, or if any criminal prosecutions will ever be launched. However, I think it's too breezy for software industry analysts to brush this case off as a low-risk threat. If Oracle’s alleged facts are true, this isn't business-as-usual; instead, this would constitute illegal marketplace behavior, with potentially severe consequences for the business generally and the decision-makers individually.
I have additional edgy things to say about this case in this interview. Other resources:
* WSJ Law Blog
* BusinessWeek article pitching this lawsuit as just bare-knuckle competition between giants
* A collection of industry analysts' comments
April 07, 2007
AFP v. Google Settles
By Eric Goldman
The Agence France-Presse v. Google lawsuit has settled. See news reports: AP, CNET News.com. Like it did with the Associated Press, Google struck a licensing deal for AFP content, including indexing it in Google News. Financial terms weren't disclosed, but I would shave my head if money wasn't moving from Google to AFP. Two observations:
1) This settlement is another example of a more content-owner-friendly Google striking deals with content owners rather than intransigently fighting them. We've seen the new warm-n-fuzzy Google working with content owners on YouTube, plus the AP deal last year, and now the AFP settlement. I expect Google to show even more conciliation with content owners in the future, not less.
2) Historically, Google's margins have been ridiculously high. If I'm right and Google is paying off both AP and AFP, I think Google's margins will decline over time as it has to pay more for content. See, e.g., Zell's announcement that he wants Google to pay up for indexing Tribune content.
More comments about the case at InfoWorld.
Previous blog coverage:
* Initial Assessment of AFP v. Google (March 2005)
* AFP v. Google Update (March 2005)
* An AFP Licensee Tells His Story of Being Kicked Out of Google News (March 2005)
* Answer in AFP v. Google (July 2005)
* Google Loses Belgium Copyright Case--Google v. Copiepresse (February 2007)
When Congress Giveth, is the Dormant Commerce Clause Taken Away?--Free Speech Coalition v. Shurtleff
by Ethan Ackerman
Free Speech Coalition, Inc. v. Shurtleff, 2:05CV949DAK (D. Utah March 23, 2007)
Why do courts seem eager to use CAN-SPAM's preemption language to give state email laws a free pass from the Dormant Commerce Clause?
Utah's courts and legislatures are earning some scrutiny lately. Seems like the general meme in Salt Lake City is to write laws that pretty clearly reach outside the Wasatch and Cottonwood valleys and touch stuff all over the internets. While some lawyers (like the state legislature's drafting counsel) are trying to reign things in, a recent Utah federal District Court ruling seems to be whipping the horses on.
The Free Speech Coalition, a trade association of "adult" publishers and marketers, has been tackling Utah's Child Protection Registry on several legal fronts for some time, and the challenges have received some coverage (see, e.g., here and here). A recent ruling in the case upholding the Utah act on several fronts has enough interesting tidbits to merit several entries, so this entry will focus on just one - the rather unique results from the intersection of preemption doctrines and the CAN-SPAM Act's preemption safe harbors.
The Harbor that Swallowed the Doctrine?
The District Court denied the Free Speech Coalition's request for a preliminary injunction of the Utah act. Contrary to the Coalition's assertions, the judge said the Utah act did not violate the Dormant Commerce Clause because it was authorized by Congress in the CAN-SPAM act. Several amici supported this argument, including the US Department of Justice and (perhaps somewhat self-interestedly) unspam, the for-profit operator of the Utah registry.
We’ve covered the Dormant Commerce Clause here before. Briefly, the Dormant Commerce Clause (DCC) constitutionally constrains some state laws, but it does not apply when federal law affirmatively authorizes those state laws. The rationale? Congress, not the states, can regulate interstate commerce - but Congress can, if it wants to, delegate its power to the states to do what they couldn't otherwise do.
Judge Kimball says that’s what happened here. Due to CAN-SPAM, he said that "Congress has expressly allowed states to regulate commercial email" by choosing not to preempt some types of state laws related to email. By creating safe harbors for some state laws, Congress effectively immunized them from DCC challenges. Judge Kimball bolstered this conclusion by citing to the similar holding in Beyond Systems v. Keynetics, which also rejected a DCC challenge against a Maryland anti-spam law on the grounds that it was covered by CAN-SPAM's safe harbor.
So why the trend away from applying the DCC? Prior to CAN-SPAM, a court had to evaluate whether a state law impermissibly burdened out-of-state commerce, and no medium is more interstate—and even international—than the Internet. Certainly the judge wasn't declining to apply a rigorous DCC analysis because it was too much work - Judge Kimball did the analysis in a belt-and-suspenders alternate holding. What magical flick of Congress' pen all of the sudden relieves a judge from evaluating the Constitutionality of a state law?
The judges in both Beyond Systems and Free Speech Coalition may be correct that state regulation explicitly authorized by federal law isn't susceptible to a DCC challenge, but it's a big jump to say that this Utah law was explicitly authorized by CAN-SPAM. To reach this conclusion, a court must pile several suspect conclusions on top of each other.
Strictly speaking, Congress didn't explicitly authorize either the Utah or Maryland acts. Instead, it referred to a class of anti-spam laws generally in the CAN-SPAM act. That's not a big deal. Congress can’t easily enumerate every unpreempted state law, and it’s impractical to have Congress update the safe harbor every time a new state law is passed or an existing one is amended. But that's also the Achilles heel of preempting state laws by general reference - we can't really tell if Congress intended to preserve a state law that wasn't even written, or was subsequently amended, when Congress drafted the safe harbor. Thus, in my opinion, the prudent approach is to construe the authorization very narrowly.
In this case, the Utah law only arguably fits into a CAN-SPAM safe harbor. CAN-SPAM has two relevant safe harbors for (1) anti-deception email-only laws, and (2) non-email-specific computer crime laws. He said the Utah act may not fit in the first category but fit into the second category of computer crime laws.
However, to reach this conclusion, the Judge had to wrestle with an internal contradiction: the Utah Act had to relate to spam to qualify for CAN-SPAM’s preemption from DCC analysis, yet the court had to simultaneously say that the law was not email-specific enough that it qualified for the non-email-specific safe harbor.
The likely analytical error here is that CAN-SPAM's preemption safe harbor intentionally sweeps wide beyond email - it's trying not to preempt any laws that don't relate to email. But that same breadth shouldn't act as a DCC shield too. Congress wasn't actually affirmatively blessing certain state email laws. It was merely acknowledging that some state laws might apply to lots of behavior (including email) that Congress would have no problem with. By rough analogy, a federal insurance law might preempt some, but not all, state insurance laws. Unrelated state insurance fraud laws would almost certainly NOT be preempted by the federal law, but that shouldn't excuse those state fraud laws from a DCC analysis.
Eric's Comment: I just want to reinforce one point in Ethan's analysis. Based on a straight reading of the CAN-SPAM statute, it is fairly clear to me that CAN-SPAM's exclusion for state computer crime laws meant only medium-neutral laws. Thus, a state email-only crime--like Utah's law here--should be preempted by CAN-SPAM. Otherwise, as Ethan points out, the exception effectively eliminates all preemptive effect of CAN-SPAM because states can freely enact any anti-spam laws--even those that conflict with CAN-SPAM or with other states' laws-- so long as the states attach criminal sanctions to the restrictions. It's possible Congress just screwed this up massively, but for now, my hypothesis is that Congress drafted the preemption language properly and it's this judge who screwed up.
April 05, 2007
Google AdWords Contract Upheld (Again)--Feldman v. Google
By Eric Goldman
Feldman v. Google, Inc., 2007 WL 966011 (E.D. Pa. March 29, 2007)
Yet another click fraud lawsuit, this time involving one of the 556 plaintiffs that opted out of the Google click fraud settlement. In my prior post, I predicted a lot of chicken-scratch litigation from those opt-outs. Here's one!
In this case, a law firm advertised via Google AdWords and allegedly was click frauded. The lawyer then sued (on behalf of his law firm) Google for click fraud in Pennsylvania. Google defended based on its AdWords contract, which has a mandatory venue provision specifying that all lawsuits shall be brought in California. We saw virtually identical facts in the initial Person v. Google case, which also involved the AdWords contract (though that lawsuit was brought in NY). The result was the same in both cases--each time, the court upheld the AdWords contract's mandatory venue clause and transferred the case to California.
Snarky aside: There's one more fact in common between Person and Feldman--both of them were lawyers bringing the lawsuit on their own behalf (pro se). Indeed, I can think of two more pro se lawyer v. Google lawsuits, the Field and Bradley cases. In each of these four cases, the lawyer representing himself got his clock cleaned by the court. Maybe this confirms the old maxim that "A person who represents himself has a fool for a client." At minimum, these pro se lawyers aren't bringing honor to themselves or our profession.
Mechanically, Google's contract formation process is bullet-proof. As the court describes:
To open an AdWords account, an advertiser had to have gone through a series of steps in an online sign-up process. (Hsu Decl. ¶ 3.) To activate the AdWords account, the advertiser had to have visited his account page, where he was shown the AdWords contract. (Hsu Decl. ¶ 4.)
Toward the top of the page displaying the AdWords contract, a notice in bold print appeared and stated, “Carefully read the following terms and conditions. If you agree with these terms, indicate your assent below.” (Hsu Decl. ¶ 4.) The terms and conditions were offered in a window, with a scroll bar that allowed the advertiser to scroll down and read the entire contract. The contract itself included the pre-amble and seven paragraphs, in twelve-point font. The contract's pre-amble, the first paragraph, and part of the second paragraph were clearly visible before scrolling down to read the rest of the contract. The preamble, visible at first impression, stated that consent to the terms listed in the Agreement constituted a binding agreement with Google. A link to a printer-friendly version of the contract was offered at the top of the contract window for the advertiser who would rather read the contract printed on paper or view it on a full-screen instead of scrolling down the window. (Hsu Decl. ¶ 5.)
At the bottom of the webpage, viewable without scrolling down, was a box and the words, “Yes, I agree to the above terms and conditions.” (Hsu Decl. ¶ 4.) The advertiser had to have clicked on this box in order to proceed to the next step. (Hsu Decl. ¶ 6.) If the advertiser did not click on “Yes, I agree ...” and instead tried to click the “Continue” button at the bottom of the webpage, the advertiser would have been returned to the same page and could not advance to the next step. If the advertiser did not agree to the AdWords contract, he could not activate his account, place any ads, or incur any charges. Plaintiff had an account activated. He placed ads and charges were incurred.
As I teach in my Cyberspace Law class, the very best online contracts are "mandatory non-leaky clickthrough" agreements. Like this one.
To get around this, the lawyer claims he was ignorant of the mandatory venue clause because he didn't read the contract. Hmm...a lawyer entering into a contract that he didn't read. Even if the contracting party weren't a lawyer, this is a pathetic argument. Every lawyer learns very, very early in their first year Contracts course that a party is bound to contract terms they assent to, even if they chose not to read the terms.
The court also slams down the plaintiff's other attacks on the contract:
* the contract didn't contain a definite price. However, the contract contained the exact formula for computing the price.
* procedural unconscionability. The court rejects this because the "Plaintiff was a sophisticated purchaser, was not in any way pressured to agree to the AdWords Agreement, was capable of understanding the Agreement's terms, consented to them, and could have rejected the Agreement with impunity."
* substantive unconscionability. The court finds many of the contract terms reasonable.
This case is a nice win for Google for two reasons. First, by upholding the mandatory venue clause, it should inhibit AdWords advertisers from suing Google all over the country. Therefore, all lawsuits will have to be in Google's home court, which raises the costs of lawsuits for most plaintiffs and gives Google some other home-court advantages. Second, by holding that this plaintiff is bound by the AdWords contract and those terms aren't substantively unconscionable, Google can now invoke its risk management clauses (like the warranty disclaimers, limits of liability, etc.) to cut the economic heart out of the click fraud claim.
April 04, 2007
Regulating "Stealth" Marketing
By Eric Goldman
I’ve never understood the legal distinctions between “editorial content” and “marketing.” To me, content is content. Certainly, content can have problematic/harmful attributes, such as false content. And often I want to know more about the identity of the content’s source so I can identify potential biases and adjust my credibility assessment of the content. However, issues such as falsity and bias can apply to editorial content and marketing equally. For example, I am just as susceptible to newspaper editors trying to manipulate my opinion as I am a marketer trying to extract some cash out of my wallet.
Having said that, I think there’s a pervasive assumption that marketing is more susceptible to bad behavior than editorial content, if for no other reason than marketers can get a financial payoff for their manipulations. And these assumptions have led to the development of an extensive regulatory framework governing marketing, ranging from substantive restrictions to limits on the medium used to deliver it. This regulatory structure vastly exceeds the regulation of editorial content.
This editorial content/marketing dichotomy has been under substantial pressure for some time. Some of this reflects the inherent illogic of the dichotomy, but technological developments have exacerbated the problems. The rise of the Internet has empowered consumers to speak up about marketplace offerings, and this content can be enormously valuable in shaping consumer perceptions and increasing producer accountability. At the same time, profit-seeking marketers have sought to surreptitiously steer consumer opinions about their offerings, leading to phenomena like shill reviews and marketer-financed blog buzz.
Ellen Goodman discussed this problem in a recent Texas Law Review article called "Stealth Marketing and Editorial Integrity." She thinks that shill Internet editorial content harms social discourse by reducing consumers’ abilities to trust content. Therefore, she favors extending sponsorship disclosure laws (like the anti-payola laws) to the Internet. (On that front, see the recent EU directive to criminalize "falsely representing oneself as a consumer.")
Stealth Marketing and Editorial Integrity is the first article in the legal literature to address the normative implications of covert marketing in mass media. For business, technological, and cultural reasons, advertisers and propagandists are increasingly using editors to pass off promotional messages as editorial content. This integration of sponsorship allows marketers to cut through communications clutter and audience resistance to marketing. In this way, the practices of payola, product placement, and sponsored journalism are proliferating and spreading into newer media forms like blogs and video games. A federal sponsorship disclosure law has proscribed these practices in broadcasting for nearly a century. Despite high-profile recent controversies about the practices, the legal literature is devoid of any systematic analysis of the problem that stealth marketing presents or the values that sponsorship disclosure might serve, whether in broadcasting or other media. This Article fills that void by providing a normative theory of sponsorship disclosure law informed by the First Amendment, bribery law, and information theory more generally.
For another exposition of her theory, see Rebecca’s summary of Ellen Goodman’s presentation at a conference.
The Texas Law Review recently launched an online counterpart called See Also, and the first installment published three critiques of Ellen's article, including my critique entitled "Stealth Risks of Regulating Stealth Marketing." My abstract:
In this response piece, Professor Goldman explores the potential adverse consequences of Professor Goodman’s proposal for sponsorship disclosure laws. More specifically, Goldman argues that any deliberation on such disclosure laws must consider: (i) why consumers desire to know the source of content; (ii) whether consumer distrust of marketing wrongly affects consumers’ evaluation of content; and (iii) the adverse effects of “noisy” disclosures.
Check out my entire comment, which does an efficient job laying out my main gripes with regulatory efforts to distinguish editorial content and marketing. (It’s relatively brief—about 2,000 words).
To keep that piece brief, I omitted two other critiques:
1) Ellen's article doesn't try to define "stealth marketing" with any precision. This gets her into trouble when she talks about blogs, where she struggles to define the border cases. This gets back to my first main critique in the posted response--why do we care about the distinction between editorial content and marketing? Without having a clearer understanding of that, there's little chance that we can adequately resolve the border cases.
2) Sponsorship disclosure laws have been on the books for four decades. Yet, I haven't seen any academic literature analyzing their efficacy. Is this right? This well-known and popular law has never been studied by academics? If it has been studied, what do those studies reveal about the efficacy of sponsorship disclosure--does it accomplish its goals? As my critique indicates, we have plenty of evidence that people say they want to know when content is marketing, yet there is also some evidence that such disclosures degrade consumer decision-making. So I'd love to see a recap of what we've learned in the past 40+ years. From my perspective, it seems odd to advocate extending the model without any evidence that the existing model actually works.
April 03, 2007
Utah Bans Keyword Advertising [Updated]
By Eric Goldman
Utah SB 236 (the "Trademark Protection Act"), enacted March 19, 2007
Legislators enact stupid laws all of the time, but some laws transcend mere stupidity and produce a single 3 letter response: WTF? And no legislature has passed more WTF Internet laws than Utah's. Consider this track record:
* in 1995, Utah enacted the nation's first digital signature legislation designed to spur PKI-based digital signatures. But no one cared, and no company ever qualified for the statutory safe harbor. Completely unused, last year Utah repealed the law entirely after 11 years of futility.
* In 2005, Utah passed a law requiring Internet access providers to allow Utah's AG nee porn czar to designate porn websites as off-limits in Utah. Utah had to repeal some of that law, but litigation over the remainder is ongoing.
* Utah recently enacted a "don't email the kids" registry. Putting aside the major problems of state-based email laws (i.e., mapping geographic-based laws onto a borderless email infrastructure), email is much more suited to client-side filtering than centralized do-not-contact registries, and there's always the risk of bad actors getting their hands on the database of kids' email addresses. As a result, this law is such a bad idea that even the consumer protection-oriented FTC advocated against it.
Based on this list alone, I think it's safe to say that Utah has an unrivaled track record of enacting dumb, regressive, unproductive Internet laws. But Utah's 3 year battle against keyword advertising represents the strongest support for this assertion.
In 2004, Utah enacted the Spyware Control Act, a completely misguided (and misnomered) law designed to protect a few noisy Utah trademark owners with weak trademarks (such as 1-800 Contacts and Overstock.com) from legitimate competition via adware. In the process, the law took technology out of consumers' hands--even if consumers wanted and valued the technology. I have previously deconstructed in great detail why this law was terrible policy.
Unfortunately, our system doesn't have good checks/balances against dumb laws other than voting the politicos out (hey, Utah readers--hint hint). Fortunately, the initial implementation of the Utah Spyware Control Act was so grossly and obviously unconstitutional that a judge had no problem quickly enjoining the law in summer 2004.
Recognizing the futility of defending that law, Utah abandoned it and amended the Spyware Control Act in 2005 to merge it with trademark law. These amendments effectively eviscerated the law because, as amended, the law required plaintiffs to establish that keyword advertising via adware made a trademark use in commerce. This legal proposition was soundly rejected in the Second Circuit's subsequent holding in 1-800 Contacts v. WhenU. After that ruling (even though it wasn't binding on Utah courts), it's relatively clear that the post-2005 Spyware Control Act failed.
Apparently undeterred by its first two misfires with the Spyware Control Act, Utah has tried to enact regressive anti-consumer legislation for a third time. This time, they've stopped messing around with adware vendors. Instead, they have made a frontal assault on all keyword advertising across-the-board. So this law now appears to cover anyone selling keyword ads, including every major search engines (including Google), many adware vendors, and plenty of other e-commerce sites (eBay, Amazon, etc.).
Specifically, the law creates a new intellectual property right called an "electronic registration mark," defined as a "word, term, or name that represents a business, goods, or a service." This definition may be broad enough to protect domain names even if the domain names are otherwise generic or unprotectable under TM law. Owners of eligible words can register the terms in a new registry by paying a nominal fee.
Once registered, an infringement occurs if another person "uses an electronic registration mark to cause the delivery or display of an advertisement for a business, goods, or a service: (i) of the same class, as defined in Section 70-3a-308, other than the business, goods, or service of the registrant of the electronic registration mark; or (ii) if that advertisement is likely to cause confusion between the business, goods, or service of the registrant of the electronic registration mark and the business, goods, or service advertised."
I read this law to restrict all competitive ad buys of registered terms, even if the advertiser is engaged in comparative advertising that would be completely permissible under existing trademark law and not confusing to any consumer. (Interestingly, the law apparently excludes ad buys by affiliates unless their ad buy causes confusion.) Both the advertiser and the ad vendor are on the hook for an infringement.
To try to limit the law's effect to just Utah, the law only applies if the ad is displayed in Utah or the advertiser or keyword vendor is located in Utah. This caveat tries to overcome the obvious dormant commerce clause problems with this law. Utah, of course, is familiar with this problem given that the first version of the Spyware Control Act was struck down on DCC grounds.
But does this qualifier save the law? The practical reality is that every advertiser, wherever they are located, would have to check Utah's registry before buying keywords that might contain a trademark of a competitor, either because the competitor might be located in Utah or the competitor might have a registration nonetheless and the ads will be displayed to Utah residents (there's no way to buy keyword ads that exclude delivery to Utah residents). So I'm 100% convinced that this law has an extraterritorial effect.
However, I've made the same argument about state do-not-spam registries (where a sender based outside Utah must check the Utah registry before sending) and other state anti-spam laws, yet most of those laws have survived a DCC challenge--including a very recent DCC challenge to Utah's don't-spam-the-kids registry (See Free Speech Coalition v. Shurtleff, 2:05CV949DAK (D. Utah March 23, 2007)). Despite that, Utah's general counsel informed the legislature that the law probably violated the DCC, and I can't imagine judges won't find that compelling. Further, there are other grounds for a challenge here, including the First Amendment and other types of preemption. So I'm reasonably confident that the law ultimately will be struck down on some basis when challenged, although plenty of resources will be needlessly spent in the process.
Irrespective of the legal analysis, I'd be remiss if I didn't say what we're all thinking: this law is terrible policy created by a legislature out of control. We've learned over the last 15 years that keywords are a uniquely empowering tool to enable consumers to express their interests more accurately, concisely and cheaply than other alternatives, which in turn enables intermediaries like search engines to cater to their informational interests. The result is lower search costs for consumers, which in turn creates big social welfare payoffs by making more socially beneficial matches between consumers and producers. So as a matter of social policy, we should be encouraging the use of keywords, not banning it (see my extended support for this argument here (and, to a lesser extent, here)).
UPDATE: Whoops, I can't believe I forgot to mention this. On top of the major DCC problems with the law, there's a very good argument that search engine/online intermediary liability under this law is preempted by 47 USC 230. Last week, the Ninth Circuit in Perfect 10 v. CCBill held that 47 USC 230 preempted all state IP claims, including state TM laws. Online intermediaries can argue that advertisers select the keywords and provide the ads, meaning those items are "provided by another information content provider," in which case online intermediaries should not be liable for that content. This argument won't help advertisers themselves, so the law still creates plenty of friction for online advertising and could still hurt online intermediaries by suppressing demand for their ad inventory. However, if other courts buy the Ninth Circuit's reading of 47 USC 230, any frontal assault on Google or adware companies using this law might very well fail.
UPDATE 2: In response to this post and others, Sen. Eastman has blogged an explanation/defense of the law. I link to his post and provide more commentary in this post.