Ad Networks Can’t Get 47 USC 230 Defense on Motion to Dismiss–Swift v. Zynga

By Eric Goldman

Swift v. Zynga Game Network, Inc., 2010 U.S. Dist. LEXIS 117355 (N.D. Cal. Nov. 3, 2010). Wendy Davis’ story on the complaint filing.

I have repeatedly observed that the Ninth Circuit’s Roommates.com en banc decision has not been the jurisprudential disaster we feared. Indeed, the strong majority of cases citing it have done so in favor of the defense. The rare exceptions include the Accusearch case (which involved really bad facts) and the NPS v. StubHub case (a state court motion to dismiss).

We can add this ruling to the list of Roommates.com pro-plaintiff citations, and in my opinion it’s the worst of the bunch because it looks past the narrow holding of Roommates.com and instead taps into the messy discussion that preceded it. My hope is that this decision is an aberration and therefore will not be influential on other courts. Unfortunately, there remains the constant risk (as yet unrealized) that the Roommates.com virus will spread in problematic ways, and this case might be the leading edge of it.

I find this case so troubling because it’s factually very similar to another post-Roommates.com case from the same district (ND Cal), Goddard v. Google, which in a thoughtful opinion cited Roommates.com in favor of the defense. Both this case and the Goddard case involved advertisements that offered some “bait” to consumers that caused responding consumers to be (allegedly) unwittingly enrolled in subscription services that had recurring charges. In the Goddard case, the bait was ringtones. In this case, the bait was “special offers” for goods/services that, when obtained, would also give the consumer some virtual currency for Zynga games.

Both Google and Zynga acted as ad networks in distributing the third party advertiser’s ads. In this case, Adknowledge apparently provided support services to Zynga’s ad network, although the complaint (and the opinion) aren’t crystal clear about those services. In general, there’s no question that 230 protects ad networks from liability for third party advertisements, even if the ads are “illegal.” The Goddard case reached this conclusion regarding Google’s ad network, and it should have worked for Zynga and Adknowledge too.

It doesn’t. The court starts out by discussing the applicability of Roommates.com. Unfortunately, this is the first time a judge has really tripped over by the murky discussion in the majority opinion about what it means to be a content “developer.” The court says:

Applying Rommates.Com [sic] here, the Court cannot determine at this juncture, based on the pleadings, whether Zynga is entitled to immunity under the CDA. Rather, the FAC alleges facts, which, if proven, could support the conclusion that Zynga is responsible, in whole or in part, for creating or developing the special offers at issue. Fundamentally, Plaintiff alleges that the special offers are desirable to users because they provide free virtual currency to be used in Zynga games. FAC ¶¶ 6, 8. In turn, Zynga is alleged to encourage acquisition of the virtual currency by designing their games to become more enjoyable as users obtain more virtual currency. Id. ¶¶ 3, 5. As noted by Plaintiff in her opposition, the lure of virtual currency is the most important “content” within the special offer because, without it, it is unlikely any user would ever participate in the offers. Additionally, Plaintiff alleges that Zynga is responsible for the design, layout, and format of the special offers, and the special offers appear directly within Zynga’s games. Id. ¶¶ 12, 13, 33, 36, 37. Moreover, Plaintiff has alleged Zynga’s “material contribution” to the alleged unlawful activity by asserting that Zynga designed its games to intentionally create the demand for the virtual currency offered in those games, and then used this demand to lure consumers into the allegedly fraudulent transactions. Id. ¶¶ 4-6, 8-9.

The court distinguishes the Goddard case because:

Plaintiff has not alleged that Zynga is a “neutral” website that merely allows third parties to post advertisements. Instead, Plaintiff asserts that Zynga is a direct participant in the fraudulent transactions that are the subject of this case, as outlined above. Therefore, at this stage, Zynga’s motion to dismiss based on CDA immunity is denied.

From my perspective, the court does not carefully distinguish between an ad network’s economic interest (which, it’s clear from the many apropos cases, is immaterial to the 47 USC 230 analysis) and an ad network’s substantive contribution to the offending content. (Even helping create the ad copy isn’t enough of a contribution to trump 230; see, e.g., Ramey v. Darkside Productions). Of course it’s in Zynga’s economic interest to give consumers reasons to want its virtual currency, and of course Zynga wants to give consumers a variety of ways to obtain that currency; that’s no different from the fact that it was in Google’s economic interest to get more advertisers and encourage them to spend more money on advertising. In both cases, however, the ad network could not govern the allegedly fraudulent conduct of rogue advertisers. Accordingly, it’s improper under 230 to extend the ad network’s financial interest in advertising to create liability for the advertiser’s rogue activity. I believe that’s what the court did here.

The court’s treatment of Adknowledge’s 230 defense is even less satisfying:

it is unclear from Plaintiff’s allegations whether Adknowledge is an “interactive computer service provider,” as that term is defined by the CDA. It is also unclear whether Adknowledge falls under the “information content provider” exception to CDA immunity. Indeed, whether Adknowledge qualifies for immunity under the CDA is a fact-based inquiry. As alleged, Adknowledge is described simply as an “aggregator” that solicits advertisements from third parties and then facilitates transactions between those parties and Zynga. FAC ¶¶ 6-8. Given the limited nature of a Rule 12(b)(6) challenge, the Court cannot determine, at this stage, whether Adknowledge is entitled to CDA immunity. It would be improper to resolve this issue on the pleadings and the limited record presented. Adknowledge’s motion to dismiss based on CDA immunity is denied.

Consistent with this, the court then notes that 230 is an affirmative defense improper for 12b6 adjudication, citing the 2008 district court opinion in Perfect 10 v. Google. Unfortunately, the court ignores that the Ninth Circuit *subsequently* tried to say the same thing in the Barnes case and then withdrew that part of the opinion, nor does the court engage the dozens of other cases saying that 230 is appropriate for a 12b6 motion to dismiss.

The discussion about Adknowledge is unsatisfying on several other fronts. Is the court really going to say that Adknowledge isn’t a provider of interactive computer services? Good luck finding precedent to support that conclusion. More generally, one of the key advantages of a 230 immunity is that it’s NOT a “fact-based inquiry”; instead, it has been used repeatedly to prevent discovery fishing expeditions. Perhaps structurally this court is waylaid by its (mistaken IMO) belief that defendants have the burden to prove 230 rather than plaintiffs having the burden to show why 230 doesn’t apply. The court’s placement of the burden on the defense is not unprecedented, but it’s pretty rare.

Even though the complaint survived the motion to dismiss, the plaintiff will still have to establish its facts. If it can’t, the defendants might be able to plead 230 at the summary judgment or trial stages.

One possible implication of this case is that an ad-supported website seeking 230 immunity for the ads should not integrate its own business offerings (such as the site’s own virtual currency) into its advertisers’ ad campaigns. Personally I don’t think this should make a difference at all to the legal analysis, but a simple change like that might have tipped this opinion in the other direction.

Another possible implication is that ad networks might voluntarily choose to do even more to identify and terminate rogue advertisers. The ad networks have an implicit moral hazard here, as they get paid by rogue advertisers, but the long-term consumer trust degradation from rogue advertisers and the possible legal exposure make those economic gains short-term at best.