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 Completely unused, last year Utah repealed the law entirely after 11 years of futility.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.