“Initial Interest Confusion”: Compounding the Error
3 O c t o b e r 2 0 0 3
JOURNAL OF INTERNET LAW
De Minimis Confusion on the Internet: Compounding the Error of Initial Interest
By Ronald D. Coleman
At an earlier stage in the development of trademark law, courts recognized the “ancient” rule that, to find an infringement, content, not merely labels, must be evaluated to determine the likelihood of confusion vel non. Thus, in Stop the Olympic Prison v. United States Olympic Commission, the court ruled that, notwithstanding the legitimate trademark rights of the US Olympic Committee in the word “Olympics,” there was no likelihood of confusion resulting from the use of the word in the name of a group formed to protest the construction of a vast Olympic Village in upper New York state./1 Similarly, in Girl Scouts v. Personality Posters Mfg. Co.,/2 the Southern District of New York rejected the suggestion that there could possibly be any actionable confusion of the sort sufficient to suggest to a consumer, endorsement, or affiliation by the Girl Scouts arising from the publication and sale of a poster featuring a pregnant young girl scout. As the court explained, far from tending to prove confusion, the swift reporting to plaintiff by offended viewers of the “offending” use actually negated the suggestion of confusion: “[I]ndignation is not confusion. To the contrary, the indignation of those who [reported the offending use] would appear to make it clear that they feel that the Girl Scouts are being unfairly put upon, not that the Girl Scouts are the manufacturers or distributors of the object of indignation.”/3
Early in the history of domain name litigation, most courts hewed to this basic rule of trademark, recognizing that a domain name is nothing more than a label and that likelihood of confusion must be considered in the entire Web site context. Courts were reluctant to extend the grant of trademark protection, which amounts to a sort of restriction on expression, beyond the congressional mandate. This was not surprising because courts had longed held that the First Amendment counsels against staking out new territory in the trademark domain.
A leading authority for this proposition is Rogers v. Grimaldi./4 In Rogers, the Second Circuit ruled that the Lanham Act must be construed narrowly when applied to titles because of First Amendment concerns. The court agreed with plaintiff Ginger Rogers that consumers might be confused by the title of a movie called, without her permission, “Ginger and Fred.” Nonetheless, the court ruled that the First Amendment protected expression that the Lanham Act might limit and found for the defendant, writing:
Since consumers expect an ordinary product to be what the name says it is, we apply the Lanham Act with some rigor to prohibit names that misdescribe such goods. But most consumers are well aware that they cannot judge a book solely by its title any more than by its cover. We therefore need not interpret the Act to require that authors select titles that unambiguously describe what the work is about nor to preclude them from using titles that are only suggestive of some topics that the work is not about. Where a title has at least some artistic relevance to the work and is not explicitly misleading as to the content of the work, it is not false advertising under the Lanham Act./5
The presumption also was that consumers possessed and used some level of discernment and judgment, and trademarks were not treated as “in gross” property rights of mark holders. Thus, in 1998, the District Court wrote in CD Solutions, Inc. v. Tooker: “While an Internet user seeking to purchase a product of defendants may go initially to the plaintiff’s website, it is unlikely than an actual or prospective customer of defendants would be confused in its purchasing decision . . . .”/6
Other courts were, through the mid-1990s, in accord./7 Consumers were expected to distinguish between the use of a mark to draw attention, especially in an editorial or noncommercial setting, and the use of a mark to indicate source. Furthermore, courts recognized that, in the appropriate circumstances, conspicuous disclaimers or clarifications of affiliation could reduce or eliminate confusion./8 Developments in technology, however, shook these principles loose. With the faddish attention paid to domain names in the late 1990s, a number of significant decisions were issued that were meant to protect owners of trademarks from the use of those trademarks by those who did not compete with them in any meaningful sense but who disagreed with them.
The fact that this change, which would ultimately affect all of the law of trademark, took place in the Web site context is significant. The concept of initial interest confusion, discussed below, had until this point been largely ignored by courts outside of the Second Circuit. But courts were apparently flummoxed in the early years of the Internet. They simply could not fathom that the use of a trademark in a domain name of a competitor, even an ideological competitor, could not have some cure. Ultimately, Congress recognized that existing law did not provide an adequate cure and so passed the Anticybersquatting Consumer Protection Act./9 Before that became law, the courts in two important and influential domain name cases fulfilled the worst legal realist nightmare, seeking to right a perceived wrong “by any means necessary.” In the process, they did substantial damage to the integrity of the law of trademark and at the same time demonstrated the worst aspects of the still nascent law of trademark dilution.
Two of the most important cases in this area involved political and controversial speech, and for that reason they are the very emblems of the misuse of trademark law. The first, which was not officially published, was Planned Parenthood Fed’n of Am. v. Bucci.10 This decision, by one-time Clinton Attorney General nominee Judge Kimba Woods, was significant for several reasons. One was the finding, unsupported by citation or even analogy, that “an intent to harm” a “competitor” is a basis for finding commercial speech under the dilution provisions of Lanham Act. Noncommercial speech is explicitly exempt from the federal antidilution statute./11 The finding of commercial speech also implicated the consideration of not only the dilution claim but also the infringement claim because commercial speech is entitled to less protection under the First Amendment./12
The Planned Parenthood court’s reasoning in this regard was based on no legal authority whatsoever. It was contrary to the Supreme Court authority that defines commercial speech in the narrowest of terms in a Lanham Act context. Moreover, it distorted the word “commercial” beyond any recognition, making a religious or ideological opponent—in this case, to abortion, a hot-button topic—a commercial “competitor,” contrary to common sense, the statute, the Constitution, and supported only by political expediency. The Planned Parenthood approach to commercial speech is not the topic of this article, and indeed the intellectual dishonesty that it and its most prominent progeny, discussed below, represents has finally been acknowledged in a recent published opinion./13 Rather, it is relevant to this discussion because it is emblematic of the extent to which courts were prepared, in the second stage of domain name litigation, to use the law of trademark as an activist bludgeon to achieve ends that could never have been intended by Congress.
Absent the congressional grant of power to address cybersquatting directly, the Planned Parenthood court bulldozed a key element of the statutory dilution tort—that the speech be commercial—in order to achieve a political end. Not surprisingly, then, the issue of likelihood of confusion was treated with similar, if somewhat less heavy-handed, contempt. The Planned Parenthood fact pattern did present an interesting legal question: At what stage of the Internet browsing experience is confusion manifest? The Web site used by the anti-abortion activist in that case was in fact deceptive. There was clear and purposeful Web site confusion based on content. In Planned Parenthood, defendant’s page began, “WELCOME TO THE PLANNED PARENTHOOD HOME PAGE!” After scrolling down and continuing to view the site, the fact that the defendant was not Planned Parenthood became apparent. In the Southern District’s view, the fact that “consumers” were deceived, however briefly, was sufficient to find a likelihood of confusion. In order to so find, and to punish a politically unpopular “bad actor,” however, the Planned Parenthood court reached for the then relatively obscure and unsettled doctrine of initial interest confusion.
Historically, even de minimis actual confusion was entitled to little weight in an analysis of likelihood of confusion./14 Initial interest confusion appears to endorse the idea that de minimis confusion is actionable not only at the actual confusion level but also at the level of likelihood of confusion. Born in the Second Circuit, it had previously been applied, according to the Federal Circuit, when a potential purchaser is initially confused such that the senior seller may be precluded from further consideration by the buyer./15 The concept, as enunciated before the more recent trend broadening
initial interest confusion, would almost axiomatically not be applicable in an Internet context, and certainly not when no commercial interest is at stake. As seen, however, the Planned Parenthood court cobbled together just such an interest, and, in terms of the limitation of the doctrine to situations implicating an actual consumer choice, this limitation was simply sub silentio jettisoned.
Ultimately, Planned Parenthood did considerable damage to the law of trademark to right a wrong that a judge considered intolerable. The harm, however, was local and limited to an unpublished decision involving an unpopular defendant engaged in a very dubious ruse. Regrettably, however, the opinion in Planned Parenthood was relied on by a judge from the other end of the ideological spectrum in a case that has become part of all too many string citations, Jews for Jesus v. Brodsky./16 This case was the first published opinion explicitly to rely on initial interest confusion when the facts objectively suggested no possible confusion at all in a trademark infringement context.
In mid-December 1997, Steven Brodsky posted an Internet Web site that he used to express his opposition to Jews for Jesus, Inc., a non-profit Christian missionary group that targets Jews for conversion to Christianity. Mr. Brodsky, a Jew, named his Web site www.jewsforjesus.org. The content of his Web site was unequivocal in its opposition to the Christian missionary group and to the concept that Jews can be “for Jesus.” It offered users the opportunity to hyperlink, or transfer, to a Web site operated by an antimissionary organization called Outreach Judaism. In late December 1997, a truthful disclaimer of affiliation with either plaintiff or Outreach Judaism was added to the display. No solicitation of funds, no advertisement, or other message bearing in any way on commerce appeared on the site. The Web sites of both plaintiff and Outreach Judaism also were non-commercial. Neither charged a fee for access nor carried commercial advertisements. Both Web sites, operated by non-profit religious organizations, offered literature promoting their respective causes and points of view, in some cases for payment of a fee. In the cases of both the Outreach Judaism and Jews for Jesus Web sites, the sale of religious materials was incidental at best to the expressive religious messages of the respective sites.
The use of the “Jews for Jesus” trademark in Brodsky by the defendant could not have caused any but the most passing confusion because a rational consumer would encounter, either immediately or following some scrolling, rhetoric antagonistic to the trademark holder on the Web site and a conspicuous disclaimer of affiliation on the screen. Nonetheless, the court in Brodsky placed great weight on a number of affidavits from individuals sympathetic to the plaintiff claiming confusion. More likely, they were not confused at all, merely angry. On reading Brodsky’s offerings, these supposedly “confused witnesses” quickly found their way to plaintiff’s Web site, undeterred by Brodsky and more zealous than ever in their devotion to plaintiff.
Absent any other authority on which base a finding of likelihood of confusion, the Brodsky court seized on Planned Parenthood. The Brodsky court relied on the Planned Parenthood court’s novel, not to say overreaching, rationale for deeming a largely, or completely, editorial Web site as “commercial speech.” It endorsed the Planned Parenthood court’s adoption of initial interest confusion as the activist solution to the lack of congressional dispatch in approving the cybersquatting law. Subsequently, on the appeal of Brodsky, the Third Circuit paid no heed to the insistence of the court across the Hudson that “indignation is not confusion”/17 and upheld the District Court in that case without comment.
After years of bubbling through the district courts, initial interest confusion was eventually endorsed by higher courts. In Promatek Indus. Ltd. v. Equitrac Corp.,/18 the Seventh Circuit used initial interest confusion to uphold a finding of infringement. The Seventh Circuit claimed to have relied on its own earlier decision in Dorr-Oliver, Inc. v. Fluid-Quip, Inc.,/19 but in fact that case made only passing reference to the Second Circuit’s jurisprudence on confusion. In fact, Promatek did not rely on initial interest confusion (it instead focuses on post-sale confusion, a different analysis) but ultimately found that there was no likelihood of confusion in the case before it, making any statement about initial interest confusion in Dorr-Oliver dictum at most./20
Why then did the Seventh Circuit so readily rely on a doctrine that has never satisfactorily been fleshed out at the Court of Appeals level? I submit that the adoption, following Brodsky, of initial interest confusion in the Internet context (mostly in the domain name context but now, because of Promatek, in a rash of metatag cases) has caused a distortion of the law of trademark, one accelerated by the effect of the promulgation of trademark dilution claims. The effect has been a weakening of the link between trademark infringement and consumer infringement and overall a reduction in free expression and even the capacity for commercial innovation.
Professor McCarthy, whose views on the topic were relied on in early initial interest infringement cases, is (not surprisingly) aware of this issue. He acknowledges that his original metaphor—a misleading sign on the highway that beckons a driver to an Exxon station, only to deliver the driver, after turning off the main road, to a “Brand X” pump— has been criticized as inapplicable to the Internet context. Consider the metatag case Bihari v. Gross/21 as follows:
Use of the highway billboard metaphor is not the best analogy to a metatag on the Internet. The harm caused by a misleading sign on the highway is difficult to correct. In contrast, on the information superhighway, resuming one’s search for the correct website is relatively simple. With one click of the mouse and a few seconds delay, a viewer can return to the search engine’s results and resume searching for the original website./22
McCarthy’s response is that “this is a difference of degree, not of kind,” because of the “unearned advantage” being gained by the one misusing the trademark to get unearned attention in the first place./23 McCarthy’s point may be a nice one standing alone in a trademark infringement context. But in a broadband trademark dilution world, it proves too much and has given trademark owners too powerful of a weapon to silence dissent or unwished attention. By and large, the main engine of trademark enforcement on the Internet is via what Siegrun Kane calls “the Dilution Solution.” It enables owners of “famous” marks to score damages and attorneys’ fees regardless of whether any harm is done to the trademark owner or even the mark itself. Like most modern intellectual property policy, the federal dilution statute is heavily weighted toward brand owners. In Moseley v. V Secret Catalogue, Inc.,24 the Supreme Court acted somewhat to counteract this trend by requiring that actual dilution, not merely the possibility of dilution, be demonstrated by a plaintiff. It has to some extent returned the focus of the trademark law not to imagined slights or what I will call “brand indignation,” but to actual damage incurred by the trademark holder. Considering the cost to free speech that wanton enforcement of the federal Dilution Act has caused, it is time to extend this protection to the dubious doctrine of de minimis, that is to say, initial interest, confusion on the Internet, to overturn the rule of Planned Parenthood and Brodsky, and to trust consumers of ideas to think for themselves.
1. Stop the Olympic Prison v. United States Olympic Commission, 489 F. Supp. 1112 (S.D.N.Y. 1980).
2. Girl Scouts v. Personality Posters Mfg. Co., 304 F. Supp. 1228, 1231 (S.D.N.Y. 1969).
3. Id. at 1231.
4. Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989).
5. Id. at 1000 (emphasis added; citations omitted). Accord, No Fear, Inc. v. Imagine Films, 930 F. Supp. 1381 (C.D. Cal. 1995); see also, Twin Peaks Prods., Inc. v. Publications Int’l, Ltd., 996 F.2d 1366, 1378 (2d Cir. 1993) (“the District Court failed to recognize the special concerns implicated by Lanham Act claims against titles of works of artistic expression”); Silverman v. CBS, Inc., 870 F.2d 40, 49 (2d Cir.), cert. denied, 492 U.S. 907 (1989) (“in determining the outer limits of trademark protection . . . the balance of [the] risks [of confusion and suppression of expression] is relevant and in some cases may tip the scales against trademark protection”).
6. CD Solutions, Inc. v. Tooker, 15 F. Supp. 2d 986, 989 (D. Ore. 1998).
7. See Data Concepts, 150 F.3d 620 623-624 (6th Cir. 1998); Interstellar Starship Svcs., Ltd. v. Epix, Inc., 983 F. Supp. 1331 (D. Ore. 1997); Teletech, 977 F. Supp. 1407, 1414 (C.D. Cal. 1997).
8. 800 Spirits Inc. v. Liquor by Wire, 14 F. Supp. 2d 675, 681 (D.N.J. 1998). Accord, Soltex Polymer Corp. v. Fortex Indus., Inc., 832 F.2d 1325, 1330 (2d Cir. 1987); Roho, Inc. v. Marquis, 902 F.2d 356, 361 (5th Cir. 1990); Better Bus. Bureau of Metro. Houston, Inc. v. Medical Dirs., Inc., 681 F.2d 397, 405 (5th Cir. 1982).
9. 15 U.S.C. §§ 1114(2)(D), 1116(a) and (c), 1117(d), 1125(d), and 1129.
10. Planned Parenthood Fed’n of Am. v. Bucci, 42 U.S.P.Q.2d 1430, (S.D.N.Y.), aff’d unpublished table decision, 152 F.3d 920 (2d Cir.), cert. denied, 119 S. Ct. 90 (1998).
11. 15 U.S.C.A. § 1125(c)(4): “The following shall not be actionable under this section . . . (B) noncommercial use of a mark.
12. Planned Parenthood Fed’n of Amer. v. Bucci, 42 U.S.P.Q.2d 1430, 1438 (S.D.N.Y. 1998).
13. Ford Motor Co. v. 2600 Enterprises, 177 F. Supp. 2d 661 (E.D. Mich. 2001). Wrote the court:
The essence of the Internet is that sites are connected to facilitate access to information. Including linked sites as grounds for finding commercial use or dilution would extend the statute far beyond its intended purpose of protecting trademark owners from use that have the effect of “lessening . . . the capacity of a famous mark to identify and distinguish goods or services.” Bally Total Fitness Holding Corp. v. Faber, 29 F. Supp. 2d 1161 (C.D. Cal.1998). This court does not believe that Congress intended the FTDA to be used by trademark holders as a tool for eliminating Internet links that, in the trademark holder’s subjective view, somehow disparage its trademark. Trademark law does not permit Plaintiff to enjoin persons from linking to its homepage simply because it does not like the domain name or other content of the linking webpage.
14. Scott Paper, 589 F.2d 1230 (3d Cir. 1978); Alpha Indus., Inc. v. Alpha Steel Tube & Shapes, Inc., 616 F.2d 440, 445 (9th Cir. 1980); AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 352 (9th Cir. 1979); First Keystone Fed. Sav. Bank v. First Keystone Mortgage, Inc., 923 F. Supp. 693, 706 (E.D. Pa. 1996).
15. Weiss Assoc., Inc. v. HRL Assoc., Inc., 902 F.2d 1546 (Fed. Cir. 1990).
16. Jews for Jesus v. Brodsky, 993 F. Supp. 282, 311 (D.N.J.), aff’d without opinion, 159 F.3d 1351 (3d Cir. 1998). The author represented defendant Brodsky in this litigation, which ended in a settlement whose terms are confidential.
17. Girl Scouts, 304 F. Supp. at 1231.
18. Promatek Indus. Ltd. v. Equitrac Corp., 63 U.S.P.Q.2d 2018 (7th Cir. 2002).
19. Dorr-Oliver, Inc. v. Fluid-Quip, Inc., 94 F.3d 376, 382 (7th Cir. 1996).
20. Thus it is not surprising that this decision was not cited by the court, not the parties, in Jews for Jesus.
21. Bihari v. Gross, 119 F. Supp. 2d 309 (S.D.N.Y. 2000).
22. Id. at 309, n.15.
23. 4 McCarthy on Trademarks § 25:69 at 25—158-159.
24. Moseley v. V Secret Catalogue, Inc., 2003 U.S. Sct. LEXIS 1945 (Mar. 4, 2003). JIL Oct 9/26/03 11:54 AM Page 6