Court Rejects First Amendment Challenge to CAN-SPAM Indictment — US v. Smallwood

[Post by Venkat Balasubramani, with comments from Ethan Ackerman]

US v. Smallwood, 09-CR-00249 (N.D. Tex.; July 15, 2011)

First Amendment challenges to spam statutes are long shots at best, with Jaynes v. Virginia being the big exception. In this case, Smallwood was charged with a variety of criminal acts, including violations of CAN-SPAM’s criminal provisions (18 U.S.C. 1037(a)(2) and (b)(2)(C)). The statute is aimed at anyone who:

uses a protected computer to relay or retransmit multiple commercial electronic mail messages, with the intent to deceive or mislead recipients, or any Internet access service, as to the origin of such messages.

Subsection (C) kicks in where there is a significant volume of messages (2,500 during any 24 hour period, 25,000 during any 30-day period, or 250,000 during any 1-year period). Smallwood argued that the statutory provisions were vague and overbroad, because it chilled “protected anonymous speech.”

The court rejects Smallwood’s arguments that the statute is overly broad. Smallwood cited Jaynes v. Virginia, a case where the Virginia Supreme Court struck down a portion of Virginia’s spam statute, but the court here distinguishes Jaynes from this case: CAN-SPAM only applies to commercial email messages. In contrast, the statute in Jaynes applies to all communications, including religious or political speech.

Smallwood’s vagueness arguments fared no better. She argued that:

she could not have known from the statute the circumstances that would cause sending multiple emails to be unlawful, because the Internet is diverse in terms of origin identification, and because multiple emailings may have multiple sources, or origins; “origin” is a broad term, defying easy definition; and there is no clear notice as to the type of conduct that might mislead or deceive someone about the origin of the emailings.

The court agrees with the government that since Smallwood is charged with something which requires a showing of intent to deceive, common sense definitions of the term “deceive” or “mislead” provides sufficient notice as to the scope of the statute.


There were two aspects of the statute that caught my attention.

First, the statute covers emails that are misleading not only to recipients, but also to any “internet access service” as to the origin. This reminds me of Kleffman v. Vonage where the California Supreme Court construed the scope of the California spam statute. (“Use of Multiple (Even Random or Garbled) Domain Names to Bypass Spam Filter Does not Violate Cal. Spam Statute.”) In that case the plaintiff argued that using multiple domain names to bypass a spam filter violated California’s spam statute. While the California Supreme Court rejected this argument, I wonder if the language in the statute (about deceiving an ISP) is aimed at something similar? Is there a legitimate vagueness argument on this point?

Second, it’s interesting that the defendant is charged with providing “SPAMmers with equipment, bandwidth, corporate infrastructure, and IP addresses, and domain names from which their customers could send SPAM with the intent to mislead recipients and Internet access services as to the messages’ origin.” Once you start penalizing people who provide bandwidth, domain names, and infrastructure, you’re pretty far afield from the people who are actually sending the emails. It looks like there’s some sort of “knowingly” limitation on this (i.e., that she supplied the assistance knowing that the customers were spammers) but it still surprised me.

Not necessarily an earthshattering ruling, but given the dearth of cases dealing with First Amendment challenges to spam statutes, an interesting data point.

Ethan’s comments:

It’s worth noting that the Court’s cursory dismissal of Smallwood’s overbreadth argument isn’t as uncontroversial as the Court suggests. The opinion dismisses Smallwood’s overbreadth argument with a quote from Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc., “the overbreadth doctrine does not apply to commercial speech.” That quote was itself a footnoted description of Central Hudson’s holding. Interestingly enough, Central Hudson doesn’t actually say anything of the sort, noting only that commercial speech is less at risk of being chilled due to overbroad laws. Supreme Court caselaw on the overbreadth doctrine as applied to commercial speech is not as simple as “not applicable.” Several Court cases describe it, like Central Hudson did, as being “seldom,” or “unlikely to be,” applicable. Bigelow v. Virginia says using the overbreadth doctrine to facially invalidate a statute restricting commercial speech is a remedy that should be used “sparingly,” but then finds the statute in question overbroad. Only Hoffman describes the overbreadth doctrine as inapplicable to commercial speech.

The Court rejects Smallwood’s vagueness challenge by adopting the Government’s argument that the intent element of the crime remedies any arguable vagueness concerns for terms like “origin.” If it explored deeper, the Court could have also pointed out that the challenged statutory terms are made rather clear by the other definitions in the Act–the defined terms “initiate,” “procure,” “sender,” “header information,” and “routine conveyance” use the contested terms, inform those terms, or provide categories that function to exclude some of those terms. It is an uphill battle to argue vagueness on a statute that has around twenty fairly specific definitions.

Previous related posts:

Internet Obscenity Conviction Requires Assessment of National Community Standards–US v. Kilbride

Still Standing? Catching Up on the Jaynes Case