With the continued growth of digital media usage and communications platforms, anonymous online postings have become more common and varied. As a result, courts, especially trial courts, continue to grapple with the best approach to deal with plaintiffs who seek to employ the courts’ substantial discovery and subpoena powers to unmask anonymous speakers.
A common situation involves a poster who, using a pseudonym and claiming to be a customer, criticizes a business. In such a case, the targeted business might seek to identify the critic not only to pursue any possible legal claim but also to publicly reveal the identity of any veiled competitor or disgruntled ex-employee. In this situation, the business might choose to file a “John Doe” lawsuit, and to do that it must serve the anonymous defendant with process. To accomplish service, a business usually files an ex parte motion asking a court to allow it to serve a subpoena or subpoenas on the defendant’s web-hosting and/or Internet service provider (ISP) seeking the anonymous poster’s Internet protocol address and user account information. This information would ideally include the address of the subscribing person (who may or may not be the same as the person who actually blogged or posted the information). The business theoretically can then use this information to make demands and/or move forward with its lawsuit.
If presented with either a plaintiff’s motion to take early discovery or enforce a subpoena seeking disclosure or a defendant’s motion to quash or seek a protective order to defeat these efforts, courts try to weigh the competing interests. On the one hand, the poster’s decision to remain anonymous (like other decisions to omit or add to content) is clearly afforded First Amendment protection—the value of anonymous works in the marketplace of ideas has been articulated by courts for decades and dates back to the unidentified pamphleteers in colonial America.1 Specifically, courts recognize that loss of such anonymity in a future lawsuit may intimidate an anonymous poster into self-censorship due to retaliation, ostracism, or loss of privacy.2 Indeed, companies aware of this chilling effect of disclosure sometimes use the powerful subpoena process to silence critics without any real interest in pursuing litigation once the subpoena “outs” the poster. On the other hand, a business or person has every right to pursue a defendant for damaging, unprotected speech3 rife on the Internet, including harassing, illegal, and defamatory online activities.
Several federal and state appellate courts have developed notice- and merit-based tests first articulated in detail in Dendrite International, Inc. v. Doe No. 3,4 briefly summarized in the next section. Several courts, however, as noted, have departed from this leaning, taking the view that a test is unnecessary because their state discovery/civil procedure rules already provide a sufficient framework.
The Dendrite/Cahill Tests
Although a large volume of literature has arisen on the appropriate procedural standard for weighing these unmasking motions, surprisingly few appellate court cases have done this with specificity.5 Perhaps this “paucity” is because discovery disputes are not usually directly appealable and mandamus review is restricted.6
In any event, one of the earliest and perhaps most well-known cases in which a procedural standard is specified is the 2001 New Jersey appellate case Dendrite.7 The Dendrite test sets out a series of showings that a plaintiff must make before the court will authorize disclosure of identifying information. The plaintiff must (1) make reasonable efforts to notify the defendant that it is seeking to identify the defendant in the lawsuit in order to allow the defendant an opportunity to oppose the request, (2) specify exactly the actionable speech, (3) reference the available legal claim, and (4) produce sufficient prima facie evidence of each element of the claim. The court must then balance the defendant’s right to anonymous speech against the plaintiff’s need for disclosure in order to determine if the petition should be granted.8
Following Dendrite, a similar procedure was set forth in Doe No. 1 v. Cahill,9 with the difference being that the Cahill test rejects the Dendrite balancing test in favor of a summary judgment showing, which, according to Cahill, itself constitutes a balancing of interests.
With each passing year, more appellate courts have adopted the Dendrite or Cahill test or some variation of one or both.10 As of 2012, higher courts in approximately eight states have chosen to follow one or the other,11 and several federal district courts have done so as well.12 Recently, a few federal appellate cases involving private parties have also addressed the use of a form of the Dendrite/Cahill test, including one in the Ninth Circuit, which held only that the adoption of Dendrite or Cahill is not clearly erroneous.13
The Civil Procedure Cases
Bucking this trend in approving Dendrite/Cahill are the civil procedure cases. One of the earlier examples is In re Cohen,14 in which a New York Supreme Court justice in a special proceeding summarily dismissed the First Amendment concerns raised by the petitioner’s application to compel preaction disclosure of an anonymous blogger, stating that identity disclosures were adequately handled by New York’s civil rules requiring the plaintiff to show a “meritorious” claim—even though no actual evidence is required. Additionally, at least one commenter has agreed. Professor Michael Vogel postulated that the practical hurdles of all of the procedural rules, including requiring court approval to identify a John Doe, helps weed out harassing claims and prevents easy destruction of anonymity.15 Few other state appellate or federal courts have explicitly taken this position, though, and it seemed for a while that In re Cohen’s view would be a fairly anomalous one.16 However, over the last few years, other courts have joined in rejecting the need to adopt any special procedure.
One of these decisions, Maxon v. Ottawa Publishing Co.,17 was issued in 2010 by an Illinois appellate court. The plaintiff petitioned the trial court for discovery against a newspaper website to allow it to identify posters who had anonymously criticized the city planning commission’s consideration of a bed and breakfast establishment proposed by the plaintiff. The newspaper opposed the petition, and the lower court dismissed the discovery attempt using a Dendrite-based test and then ruled that the statements were nonactionable opinions as a matter of law. The appellate court, however, rejected the trial court’s use of Dendrite. The appellate court said that the state’s civil procedure rules allowing for a presuit petition to identify unknown defendants, which must be verified and provide grounds for the request and a hearing on the cause of action, were sufficient to avoid a “fishing expedition” by the plaintiff.18 Because the petition was subject to an attack under the rules based on defects apparent on the face of the petition, surviving such an attack would be sufficient to protect anonymous speech. According to the court, the Dendrite balancing test and the “hypothetical motion for summary judgment” required by Cahill were not necessary.19
In April 2013, a Michigan appellate court weighed in with similar reasoning in Cooley School of Law v. Doe 1.20 Cooley dealt with a series of harshly critical remarks by an alleged former student posted on his website, which was hosted by a California-based web services company, Weebly.com. The law school sued Doe 1 for defamation and tortious interference in a Michigan trial court and then proceeded to petition a California court to subpoena Weebly.com to produce user account information. After the California court granted the subpoena, Doe 1 filed a motion in the Michigan court to quash it. The trial court, in a long analysis relying on Dendrite’s balancing test, denied Doe’s motion to quash the subpoena and declined to grant him a protective order. On appeal, the majority opinion considered the trial court’s use of the Dendrite test unnecessary because Michigan’s court rules “largely overlap” the Dendrite requirements to adequately protect the defendant’s First Amendment rights in anonymity.21 The court said that the summary disposition motion, which tests the legal basis of the complaint on the face of the pleadings alone, and protective orders were sufficient to protect the defendant’s interest in anonymity. The appellate court said that in applying Dendrite, the trial court had abused its discretion; the appellate court reversed the decision, remanding the matter back to the lower court to determine if a protective order was applicable to protect John Doe’s identity.
Maxon/Cooley Problem Areas
These cases highlight serious speech concerns about whether civil discovery rules can adequately protect the First Amendment rights of anonymous posters in the varying circumstances that may arise. These concerns, many of which were raised in the cogent dissent filed with each opinion, include inadequate standards and evidentiary showing to protect speech, inadequate notice, and lack of a balancing test.
Inadequate Standards and Evidentiary Showing to Protect Speech
Maxon’s reliance on the Illinois rules requiring a verified petition and hearing and allowing for attacks on the legal sufficiency of a plaintiff’s underlying claim in its unmasking petition is inadequate due to the limited showing required from the plaintiff. Specifically, as the dissent pointed out, the rules broadly allow consideration of what has been pleaded as “facts,” not what is actually proven. There is no assurance that a plaintiff has a supportable defamation claim. If baseless facts are initially pleaded and lead to the discovery of the poster’s identity, “the harm to anonymous speech is a fait accompli” because the court cannot later undo the prior harm of disclosure.22 The dissent, whose concerns were not allayed by the fact that Illinois is a verified fact pleading state, noted that such verified pleadings have not “eliminated frivolous lawsuits” and that “[p]laintiffs routinely plead ‘facts’ which later cannot be proven.”23
In Cooley, the majority found that a motion for “summary disposition” combined with protective orders sufficiently protects the anonymous speaker. However, the Cooley dissent, in reasoning similar to the Maxon dissent, noted that the summary disposition standard only tests the legal basis of the complaint on the pleadings alone and that the court must accept the factual allegations in the complaint as true and view them in the most favorable light of the nonmoving party. This apparently was of particular concern to the dissent in light of the fact that Michigan is a notice pleading state, which requires only minimal allegations to inform the opposing party of the claim. As the dissenting judge noted, the “relative ease” to plead a survivable claim allowed for a bare-bones showing of the plaintiff’s underlying defamation or other claim, which was insufficient to properly weigh and protect the speech interests involved.24
The Cooley and Maxon dissents recognized that the rules can allow for invasion of privacy of an innocent person, especially in the context of an ex parte proceeding, based on allegations in pleadings that are arguably vague even where they were verified and survived a motion to dismiss. To invoke the court’s use of its extraordinary unmasking powers, a higher vetting standard is needed. Indeed, courts and commenters have compared the court’s use of preservice discovery to obtaining warrants in criminal investigations, where probable cause is necessary.25
The arguable failure of the civil procedure rules to impose a strong enough standard gives rise to another related concern—that is, any test that is used must clearly illustrate what evidentiary showing is required. Even courts that have developed a separate test have applied differing and sometimes confusing showings. Here are four of them, based on increasing severity as it is traditionally understood: (1) the plaintiff must provide a good faith basis upon which to bring the underlying claim; (2) the claim must survive a motion to dismiss for failure to state a cause of action standard confined to the pleadings; (3) the plaintiff must establish a prima facie case for each element; and (4) the claim must meet Cahill’s summary judgment standard, i.e., that there is no genuine dispute of material fact that the plaintiff is entitled to judgment. Although courts and some commenters have viewed the good faith standard as the laxest and the summary judgment standard as the most rigorous, it is sometimes unclear where all of these labels actually fall on the permissive/strict spectrum. For example, if a prima facie showing is what the court decides to require, isn’t that in itself a vague standard as to what evidence is actually required? And, in fact, don’t prima facie showings vary from state to state? Additionally, as pointed out by one commenter,26 In re Cohen seems to use a relatively lax motion to dismiss standard confined to the face of the pleadings, but the court in In re Subpoena Duces Tecum to America Online, Inc.,27 applying the less strict good faith standard, actually requires additional evidence beyond the four corners of the complaint.
Considering the confusion, plaintiffs at the very least should be required to provide a “real evidentiary basis” and “adduce competent evidence . . . [that] must address all of the inferences of fact that plaintiff[s] would need to prove in order to prevail.”28 Further, courts should not just use a label but, rather, clarify what is required in terms of evidence. A California court addressing this issue found “it unnecessary and potentially confusing to attach a procedural label, whether summary judgment or motion to dismiss.”29 For example, if a prima facie standard is used, the court should specifically require an affidavit, deposition, admission, or another statement under oath beyond just a complaint, verified or not.
On the other hand, it is important to remember that although standards and evidentiary showings certainly must be clear and robustly protect at the front end against unnecessary disclosure, it is not necessary that the burden be inordinately severe so that it is unfair to the plaintiff or unreasonably taxes the resources of the court. As recognized in Cahill, a plaintiff’s burden should not include requiring evidence that cannot be expected to be in the plaintiff’s control. For example, evidence of intent, such as malice in a defamation case against a public figure, would not be required because the plaintiff likely would not have that information easily accessible at that point.30
As can be seen, determining the proper standard and showing can be a difficult and sensitive task that the civil rules do not envision and cannot replace.
The Cooley dissent raised another interesting deficiency in using the civil rules to vet unmasking petitions: that deficiency relates to notice, which was also recognized by the majority but discounted. Specifically, the problem arises when a plaintiff seeks to obtain John Doe’s identity so that extrajudicial efforts can be taken to silence the critic. In such a case, the plaintiff can simply file suit and then quickly seek discovery from the third-party web host to gain the identifying information. In the meantime, the defendant may not have notice to defend himself against the efforts (e.g., by asking for a protective order); and by the time he is notified, it is too late. The Dendrite/Cahill line of cases would at least try to prevent this by requiring the plaintiff to make “reasonable efforts” to give notice to the unknown defendant by, for example, posting an alert on the same platform on which the criticism was delivered.
The Cooley majority dismissed this notice problem as an “extreme case” and said that use of court rules would be adequate in nearly every situation to protect anonymity.31 However, the majority’s extreme case may, to the contrary, actually be—at least in the online environment—the “ordinary” fact pattern. The Cooley case itself, as pointed out by the dissent, almost involved such a pattern. Doe 1 did not receive notice from Cooley of the lawsuit or of the subpoena it obtained from Weebly.com but later learned of it from a press release that Cooley issued after it was filed. If, by happenstance, the media had not alerted the defendant, his name could have easily been identified and publicized by Cooley without the defendant having a chance to ask for a protective order.
As touched on above, requiring notice from the plaintiff is also critical in light of the fact that damage from identifying an anonymous poster is not likely to be remedied—that is, once anonymity is lost, it cannot be “un-unmasked.” Without proper notice, the John Doe proceeding can easily become a single-party event where the plaintiff is provided with important relief that the court cannot later reverse. In light of the importance of notification, leaving this task to the good will of a plaintiff or the terms of the Internet provider, as state rules would do if used, should be avoided.
Lack of a Balancing Test
From a speech analysis perspective, the civil rules’ failure to provide a Dendrite-type balancing test is unsatisfactory. The balancing process explicitly requires the court to weigh the value of keeping anonymous a particular type or source of speech against the plaintiff’s need to identify the speaker in order to proceed. For example, in a balancing analysis, the underlying cause of action might impact the level of scrutiny that a court may bring to bear on the plaintiff’s unmasking request. If the action is based on clearly commercial speech, such as product advertising (which is viewed as a particularly resilient form because it is incentivized by economic benefit), courts may deem it subject to limited First Amendment protection. In such a case, the court may give less credence to protecting John Doe’s anonymity. Recently, federal courts have suggested that this is the case in the file-sharing copyright infringement cases (which are beyond the scope of this article but contain interesting reading).34 On the other hand, in balancing the interests, courts can recognize absolutely critical anonymous speech, such as criticism of powerful public figures, government whistle-blowing, or even voices of dissidents of repressive regimes. Furthermore, some courts consider additional factors to balance, such as “whether the plaintiff truly needs the identity disclosed and whether there is an alternative way to obtain the information.”35
Although it is recognized that Cahill declined to adopt such a balancing test—finding that the summary judgment test is a balancing test and that a separate test needlessly complicates the analysis—the balancing does not need to be complicated and is useful in that it gives the court needed flexibility to appropriately weigh the varied situations that underlie unmasking petitions.
Discovery rules alone are no substitute for developing a thoroughly considered Dendrite/Cahill-type test to protect anonymous speech. As the cases percolate through the system, the challenge of the appellate courts will be to formulate clear guidance for lower courts that strikes the right balance in preserving an old and particularly useful vehicle for the public communication of ideas while recognizing a merit-based right of a plaintiff to bring online wrongdoers to justice. Courts will want to develop a test that weeds out abusive plaintiffs who seek a “heckler’s veto”—that is, plaintiffs who try to use the courts’ powerful subpoena or early discovery powers to track down their critics’ identity and then silence them extrajudicially. Although on their surface sufficient, discovery rules alone contain built-in limits that make them unsatisfactory for the complexity of many unmasking decisions. As more courts address the issue, hopefully a single standard will emerge.36 Until then, courts will need to walk a fine line in developing their own balanced and refined tests.
1. McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 341–42 (1995).
3. Beauharnais v. Illinois, 343 U.S. 250, 266 (1952).
4. 775 A.2d 756 (N.J. Super. Ct. App. Div. 2001).
5. However, lower courts do encounter the issue on a regular basis, as has been the case in Florida. See, e.g., Costello v. SuperValu, Inc., Case No. 41-2009-CA-001150 (Fla. Cir. Ct., Manatee County Aug. 30, 2010) (order); Mohlman v. Investorshub.com, Inc., Case No. 37 2012CA000040 (Fla. Cir. Ct., Leon County Mar. 1, 2012) (order).
6. In re Anonymous Online Speakers, 611 F.3d 653, 659 (9th Cir. 2010).
7. Dendrite Int’l, Inc. v. Doe No. 3, 775 A.2d 756 (N.J. Super. Ct. App. Div. 2001).
8. Id. at 760.
9. 884 A.2d 451 (Del. 2005).
10. See Paul Alan Levy, Developments in Dendrite, 14 Fla. Coastal L. Rev. 1, 12–16 (2012).
11. Krinsky v. Doe 6, 72 Cal. Rptr. 3d 231, 245–46 (Ct. App. 2008); Solers v. Doe, 977 A.2d 941, 954 (D.C. 2009); In re Does 1–10, 242 S.W.3d 805, 821 (Tex. App. 2007); Mobilisa, Inc. v. Doe 1, 170 P.3d 712, 720 (Ariz. Ct. App. 2007); In re Ind. Newspapers, Inc., 963 N.E.2d 534, 552 (Ind. Ct. App. 2012); Indep. Newspapers, Inc. v. Brodie, 966 A.2d 432, 456–57 (Md. 2009); Mortgage Specialists, Inc. v. Implode Explode Heavy Indus., Inc., 999 A.2d 184, 193 (N.H. 2010); Pilchesky v. Gatelli, 12 A.3d 430, 445 (Pa. Super. Ct. 2011).
12. See, e.g., Art of Living Found. v. Does 1–10, No. 10 CV 05022 LHK, 2011 WL 5444622 (N.D. Cal. Nov. 9, 2011); Koch Indus., Inc. v. Does 1–25, No. 2:10CV1275DAK, 2011 WL 1775765 (D. Utah May 9, 2011); Highfields Capital Mgmt. v. Doe, 385 F. Supp. 2d 969 (N.D. Cal. 2005); Fodor v. Doe, No. 3:10 CV 0798 RCJ (VPC), 2011 WL 1629572, at *3–4 (D. Nev. Apr. 27, 2011); Doe I v. Individuals, 561 F. Supp. 2d 249, 255 (D. Conn. 2008); London Sire Records, Inc. v. Doe 1, 542 F. Supp. 2d 153, 164 n.12 (D. Mass. 2008); Alvis Coatings, Inc. v. Does 1–10, No. 3L94 CV 374-H, 2004 WL 2904405, at *3 (W.D.N.C. Dec. 2, 2004); Sony Music Entm’t Inc. v. Does 1–40, 326 F. Supp. 2d 556, 563–64 (S.D.N.Y. 2004). See Levy, supra note 10, at 12–15, for a more informative and complete list.
13. In re Anonymous Online Speakers, 661 F.3d 1168 (9th Cir. 2011).
14. 887 N.Y.S.2d 424 (N.Y. Sup. Ct. 2009).
15. See, e.g., Michael S. Vogel, Unmasking “John Doe” Defendants: The Case Against Excessive Hand-Wringing over Legal Standards, 83 Or. L. Rev. 795, 855–56 (2004).
16. There have been a few other cases besides those mentioned above. Stone v. Paddock Publications, Inc., 961 N.E.2d 380, 388–90 (Ill. App. Ct. 2011), citing Maxon, found that the state rules requiring a verified complaint, specification of the defamatory statement, and notice to defendant made it unnecessary to apply the First Amendment to the petition. In Lassa v. Rongstad, 718 N.W.2d 673 (Wis. 2006), a nondigital case, the court said that Wisconsin’s detailed pleading requirements met Cahill’s First Amendment concerns.
17. 402 Ill. App. 3d 704, 929 N.E. 2d 666 (2010).
18. 402 Ill. App. 3d at 711.
19. Id. at 714–15.
20. 2013 Mich. App. LEXIS 610 (Apr. 4, 2013).
22. Maxon, 929 N.E.2d at 679 (emphasis in original).
24. Cooley, 2013 Mich. App. LEXIS 610, at *50.
25. Columbia Ins. Co. v. Seescandy.com, 185 F.R.D. 573, 579–80 (N.D. Cal. 1999).
26. See Michael R. Baumrind, Note and Comment: Protecting Online Anonymity and Preserving Reputation Through Due Process, 27 Ga. St. U. L. Rev. 757, 764–68 (2011).
27. 52 Va. Cir. 26, 37 (2000).
28. Highfields Capital Mgmt. L.P. v. Doe, 385 F. Supp. 2d 969, 975–76 (N.D. Cal. 2005).
29. Krinsky v. Doe 6, 72 Cal. Rptr. 3d 231, 244 (Ct. App. 2008); see Baumrind, supra note 26, at 769.
30. Doe No. 1 v. Cahill, 884 A.2d 451, 464 (Del. 2005).
31. Cooley Sch. of Law v. Doe 1, 2013 Mich. App. LEXIS 610 (Apr. 4, 2013).
32. 887 N.Y.S.2d at 425.
33. See Levy, supra note 10, at 28–47.
34. Sony Music Entm’t Inc. v. Does 1–40, 326 F. Supp. 2d 556 (S.D.N.Y. 2004); In re Anonymous Online Speakers, 661 F.3d 1168 (9th Cir. 2011).
35. See Baumrind, supra note 26, at 769–70; Doe v. Individuals, 561 F. Supp. 2d 249, 255 (D. Conn. 2008).
36. See Nathaniel Gleicher, John Doe Subpoenas: Toward a Consistent Legal Standard, 118 Yale L.J. 320 (2008).