holding that while there is support for both a narrow and broad reading of "exceeds authorized access," the rule of lenity requires the court to adopt a narrower interpretation in the defendant's favorSummary of this case from United States v. Nosal
Nos. 14–2710–cr 14–4396–cr.
Justin Anderson and Randall W. Jackson (Hadassa Waxman and Brooke Cucinella, of counsel), Assistant United States Attorneys fori Preet Bharara, United States Attorney for the Southern District of New York, New York, N.Y., for Appellant/Appellee. Edward S. Zas (Robert M. Baum and Julia L. Gatto, of counsel), Federal Defenders of New York, Inc., New York, N.Y., for Defendant–Appellee/Defendant–Appellant Gilberto Valle. Eugene Volokh (Hanni Fakhoury and Jamie Williams, Electronic Frontier Foundation, San Francisco, CA, on the brief), Scott & Cyan Banister First Amendment Clinic, UCLA School of Law, Los Angeles, CA, for Amici Curiae Electronic Frontier Foundation, Center for Democracy & Technology, Marion B. Brechner First Amendment Project, National Coalition Against Censorship, Pennsylvania Center for the First Amendment, and Law Professors. Stephen L. Braga, Appellate Litigation Clinic, University of Virginia School of Law, Charlottesville, VA, for Amici Curiae Frederick S. Berlin, M.D., Ph.D., and Chris Kraft, Ph.D. Hanni Fakhoury and Jamie Williams (Richard D. Willstatter, National Association of Criminal Defense Lawyers, White Plains, New York, and Harley Geiger, Center for Democracy & Technology, Washington, D.C., on the brief), Electronic Frontier Foundation, San Francisco, CA, for Amici Curiae Electronic Frontier Foundation, Center for Democracy & Technology, National Association of Criminal Defense Lawyers, and Scholars.
Justin Anderson and Randall W. Jackson (Hadassa Waxman and Brooke Cucinella, of counsel), Assistant United States Attorneys fori Preet Bharara, United States Attorney for the Southern District of New York, New York, N.Y., for Appellant/Appellee.
Edward S. Zas (Robert M. Baum and Julia L. Gatto, of counsel), Federal Defenders of New York, Inc., New York, N.Y., for Defendant–Appellee/Defendant–Appellant Gilberto Valle.
Eugene Volokh (Hanni Fakhoury and Jamie Williams, Electronic Frontier Foundation, San Francisco, CA, on the brief), Scott & Cyan Banister First Amendment Clinic, UCLA School of Law, Los Angeles, CA, for Amici Curiae Electronic Frontier Foundation, Center for Democracy & Technology, Marion B. Brechner First Amendment Project, National Coalition Against Censorship, Pennsylvania Center for the First Amendment, and Law Professors.
Stephen L. Braga, Appellate Litigation Clinic, University of Virginia School of Law, Charlottesville, VA, for Amici Curiae Frederick S. Berlin, M.D., Ph.D., and Chris Kraft, Ph.D.
Hanni Fakhoury and Jamie Williams (Richard D. Willstatter, National Association of Criminal Defense Lawyers, White Plains, New York, and Harley Geiger, Center for Democracy & Technology, Washington, D.C., on the brief), Electronic Frontier Foundation, San Francisco, CA, for Amici Curiae Electronic Frontier Foundation, Center for Democracy & Technology, National Association of Criminal Defense Lawyers, and Scholars.
BARRINGTON D. PARKER, Circuit Judge:
This is a case about the line between fantasy and criminal intent. Although it is increasingly challenging to identify that line in the Internet age, it still exists and it must be rationally discernible in order to ensure that “a person's inclinations and fantasies are his own and beyond the reach of the government.” Jacobson v. United States, 503 U.S. 540, 551–52, 112 S.Ct. 1535, 118 L.Ed.2d 174 (1992). We are loath to give the government the power to punish us for our thoughts and not our actions. Stanley v. Georgia, 394 U.S. 557, 565, 89 S.Ct. 1243, 22 L.Ed.2d 542 (1969). That includes the power to criminalize an individual's expression of sexual fantasies, no matter how perverse or disturbing. Fantasizing about committing a crime, even a crime of violence against a real person whom you know, is not a crime.
This does not mean that fantasies are harmless. To the contrary, fantasies of violence against women are both a symptom of and a contributor to a culture of exploitation, a massive social harm that demeans women. Yet we must not forget that in a free and functioning society, not every harm is meant to be addressed with the federal criminal law. Because “[t]he link between fantasy and intent is too tenuous for fantasy [alone] to be probative,” United States v. Curtin, 489 F.3d 935, 961 (9th Cir.2007) (en banc) (Kleinfeld, J., concurring), and because the remaining evidence is insufficient to prove the existence of an illegal agreement or Valle's specific intent to kidnap anyone, we affirm the district court's judgment of acquittal on the single count of conspiracy to kidnap.
In an issue of first impression that has sharply divided our sister circuits, we must also decide the meaning of “exceeds authorized access” in section 1030(a) of the Computer Fraud and Abuse Act (“CFAA”), which imposes both criminal and civil liability. 18 U.S.C. § 1030. Specifically, we must determine whether an individual “exceeds authorized access” to a computer when, with an improper purpose, he accesses a computer to obtain or alter information that he is otherwise authorized to access, or if he “exceeds authorized access” only when he obtains or alters information that he does not have authorization to access for any purpose which is located on a computer that he is otherwise authorized to access. Because we conclude that the text, statutory history, and purpose of the CFAA permit both interpretations, we are required to apply the rule of lenity and adopt the latter construction. We therefore reverse the judgment of conviction as to the CFAA count.
Gilberto Valle is a native of Forest Hills, Queens. At the time of the events giving rise to his prosecution, he was an officer in the New York City Police Department living with his wife, Kathleen Mangan, and their infant daughter in Forest Hills. Valle has no prior criminal record and there is no evidence that he ever acted violently or threateningly towards anyone.
Valle was, however, an active member of an Internet sex fetish community called Dark Fetish Network (“DFN”). He connected with individuals around the world whom he knew only by screen names such as “Moody Blues” or “Aly Kahn,” or by email addresses. Valle communicated with these individuals by email or web chat, usually in the late evening and early morning hours after his work shift. Many of his Internet communications involved the transmission of photographs of women he knew—including his wife, her colleagues from work, and some of his friends and acquaintances—to other DFN users with whom he discussed committing horrific acts of sexual violence. These “chats” consisted of gruesome and graphic descriptions of kidnapping, torturing, cooking, raping, murdering, and cannibalizing various women.
Valle's online fantasy life was, to say the least, extremely active during this period. However, there is no evidence that he ever learned the real identities of the individuals with whom he chatted, nor is there any evidence that he ever made concrete plans to meet in person or speak by telephone or web camera with any of them.
In September 2012, Mangan became concerned about Valle's late-night Internet activities after she found several disturbing images of dead women on a laptop that the couple shared. She installed spyware on the computer, which recorded each website entered by the computer's users and captured screen shots every five minutes. With the use of the spyware, Mangan found more disturbing pictures and records of websites that Valle visited. These included detailed emails and chats where Valle discussed butchering her and raping and torturing other women whom they knew. After confronting Valle about his computer use and moving out of the home with their daughter, Mangan contacted federal authorities.
Valle was subsequently arrested and charged with a single conspiracy to kidnap several of the women who were the subject of his chats. Although he had chatted with numerous individuals he met on DFN, the Government identified three alleged co-conspirators: Michael VanHise, a man from New Jersey who was known to Valle as “firstname.lastname@example.org” and “email@example.com”; an unidentified individual apparently located in Pakistan who used the screen name “Aly Khan”; and Dale Bolinger, a man in England who was known to Valle only by his screen name, “Moody Blues.” And although Valle had discussed up to one hundred different women in his chats, the indictment alleged five targets of the kidnapping conspiracy: Kathleen Mangan, his wife; Alisa Friscia, Mangan's former co-worker; Andria Noble; Kristen Ponticelli; and Kimberly Sauer, a former college classmate of Valle's who was living in the Baltimore area.
Valle was also charged with improperly accessing a government computer and obtaining information, in violation of section 1030(a)(2)(B) of the CFAA. As an NYPD officer, Valle had access to the Omnixx Force Mobile (“OFM”), a computer program that allows officers to search various restricted databases, including the federal National Crime Information Center database, which contain sensitive information about individuals such as home addresses and dates of birth. It is undisputed that the NYPD's policy, known to Valle, was that these databases could only be accessed in the course of an officer's official duties and that accessing them for personal use violated Department rules. In May 2012, he accessed the OFM and searched for Maureen Hartigan, a woman he had known since high school and had discussed kidnapping with Aly Khan. This access with no law enforcement purpose is the basis for the CFAA charge.
The Government's evidence at trial included the chats and emails between Valle and his alleged co-conspirators; testimony from several of the alleged targets of the kidnapping conspiracy, including his wife; other evidence seized from Valle's computer, including videos and images he downloaded; his search term and browser history; and excerpts from a post-arrest statement. Following a 13–day trial, the jury returned a verdict of guilty on both counts. Valle subsequently moved for a judgment of acquittal pursuant to Rule 29 or, in the alternative, for a new trial pursuant to Rule 33 on both counts.
In a thorough and thoughtful 118–page opinion, the district court (Gardephe, J.) granted Valle's Rule 29 motion with respect to the conspiracy charge. 301 F.R.D. 53 (S.D.N.Y.2014). While remaining “mindful of the jury's critical role in our legal system,” Judge Gardephe acknowledged his responsibility to ensure that the government satisfies its burden of establishing proof beyond a reasonable doubt. Id. at 80. Emphasizing “the unique circumstances of this extraordinary case,” he concluded that, notwithstanding the jury's verdict to the contrary, the prosecutors had failed to prove beyond a reasonable doubt that Valle and his alleged co-conspirators had entered into a conspiracy to kidnap or that Valle had formed the requisite specific intent to kidnap. Id. at 62, 89.
In reaching this conclusion, Judge Gardephe cited extensively to the testimony of FBI Special Agent Corey Walsh, the lead agent assigned to review and analyze Valle's emails and chats whose testimony had formed (in the court's view) the “centerpiece” of the Government's case and the “foundation” of its argument that Valle had acted with criminal intent. Id. at 83–84. Agent Walsh testified that he, along with prosecutors and other case agents, reviewed all of the emails and chats found on Valle's computer and concluded that Valle's conversations with 21 of the 24 individuals whom he “met” on DFN were “fantasy.” SA 8, 128. At the same time, the prosecution team concluded that Valle's conversations with the three alleged co-conspirators contained what they termed “elements of real crime” because they “described dates, names, and activities that you would use to conduct a real crime.” 301 F.R.D. at 65. There was no evidence that Agent Walsh or any of the other members of the prosecution team had any specialized training or experience that would render them particularly competent to distinguish between “real” and “fantasy” chats. Indeed, Agent Walsh conceded that the “fantasy role-play” chats and emails shared many of the same features as the “real” chats and emails that purportedly reflected criminal intent, including dates for planned kidnappings, conjured acts of sexual violence, prior surveillance that Valle fantasized about having conducted, and fantastical elements such as human-sized ovens and rotisseries for cooking victims. Id. at 65–66.
After an exhaustive review of the chats and emails introduced at trial, Judge Gardephe concluded that there was no discernible difference between the “real” and “fantasy” chats:
Both sets of chats involve discussions about Facebook photographs of women Valle knows; dates for planned kidnappings; prices Valle will charge for kidnapping these women; surveillance Valle has allegedly conducted of these women; the use of chloroform to incapacitate victims; acts of sexual violence that will be perpetrated on these women; and fantastical elements such as human-size ovens and rotisseries, and the construction of soundproofed basements and pulley apparatuses that will be used for purposes of torture.
Id. at 60. Accordingly, he concluded that no reasonable juror could have found beyond a reasonable doubt that the allegedly “real” chats evinced criminal intent any more than did the acknowledged “fantasy” chats. Id. at 84.
The district court further concluded that the Government's remaining evidence, including Valle's Internet search history and “real life” encounters with several of the alleged targets, was insufficient to establish either a genuine agreement to kidnap or Valle's specific intent to kidnap in light of the fantastical nature of the chats and the weakness of the remaining evidence. Id. at 90. Judge Gardephe stressed, among other things, that there was no evidence that any of the alleged conspirators ever exchanged contact information or sought to learn each other's true identities, and that the communications were episodic, with months often passing in between. Id. at 60. When dates for planned kidnappings came and went, Valle and his alleged co-conspirators would “simply begin discussing another woman as a potential target, in the same manner that a consumer of pornography might turn to a different image, photograph, or movie.” Id. at 89. They also had agreed to the impossible—kidnapping three different women in three different places spanning thousands of miles on the same day—and Valle had “provided his alleged co-conspirators with a veritable avalanche of false, fictitious, and fantastical information concerning himself and the steps he had allegedly taken to facilitate a kidnapping.” Id. at 61, 90. These facts, Judge Gardephe reasoned, were “entirely inconsistent with the notion that Valle was engaged in a genuine kidnapping conspiracy” and, on the other hand, “entirely consistent with Valle's defense that he was engaged in fantasy role-play” and that the intent of the conversations was simply “mutual fantasizing.” Id. at 60, 90. Accordingly, Judge Gardephe concluded that the Government's proof had not established Valle's guilt beyond a reasonable doubt and granted Valle's motion for a judgment of acquittal.
For many of the same reasons, Judge Gardephe conditionally granted Valle's motion for a new trial on the ground that the jury's verdict was contrary to the weight of the evidence. Id. at 104. Although the basis for his ruling was limited to the weight of the evidence, Judge Gardephe also expressed serious concern about the prosecution's trial tactics and the effect they may have had on the jury. Specifically, he questioned the propriety of the prosecution's repeated references to Valle's status as a police officer, such as arguments that “it is not ok” for someone in that position to engage in such fantasies. “Once the lies and the fantastical elements [of the chats] are stripped away,” Judge Gardephe concluded, “what is left are deeply disturbing misogynistic chats and emails written by an individual obsessed with imagining women he knows suffering horrific sex-related pain, terror, and degradation.” Id. at 61. “[I]n what was an extraordinary case involving highly inflammatory and emotional subjects,” the prosecution's questionable conduct had “raise[d] concerns” that the jury's verdict was the product of “disgust and revulsion” rather than reason and that Valle had been “held to a higher standard because of his status as a police officer.” Id. at 105–07, 109.
Finally, the district court denied Valle's motion for a judgment of acquittal as to the CFAA count. While acknowledging the existence of a “vigorous judicial debate” over the meaning of “exceeds authorized access,” the court nonetheless concluded that Valle's conduct fell “squarely within the plain language” of the statute because Valle had not been authorized “to input a query regarding Hartigan's name” without a law enforcement reason for doing so. Id. at 111, 113.
Valle was sentenced to 12 months in custody (which was principally a sentence of time served because he had already spent 20 months in pretrial detention), one year of supervised release, and a $25 special assessment. The Government has appealed the judgment of acquittal on the conspiracy count and Valle has appealed his conviction on the CFAA count.
STANDARD OF REVIEW
When reviewing a judgment of acquittal under Rule 29, we view the evidence in the light most favorable to the Government with all reasonable inferences resolved in the Government's favor. United States v. Anderson, 747 F.3d 51, 60 (2d Cir.2014). The ultimate question is not whether we believe the evidence adduced at trial established the defendant's guilt beyond a reasonable doubt, but whether any rational trier could reasonably reach that conclusion. United States v. Payton, 159 F.3d 49, 56 (2d Cir.1998) (citing Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979)). Accordingly, a jury's verdict must be upheld if any rational trier of fact could have found the essential elements of the crime had been proved beyond a reasonable doubt. United States v. Coplan, 703 F.3d 46, 62 (2d Cir.2012).
Applying this standard does not, however, mean that a reviewing court must affirm all jury verdicts. If “we are to be faithful to the constitutional requirement that no person may be convicted unless the Government has proven guilt beyond a reasonable doubt, we must take seriously our obligation to assess the record to determine ... whether a jury could reasonably find guilt beyond a reasonable doubt.” United States v. Clark, 740 F.3d 808, 811 (2d Cir.2014). This standard does not mean that if there is any evidence that arguably could support a verdict, we must affirm. In any criminal trial there is always some evidence of guilt, otherwise there could not have been a prosecution.
While we defer to a jury's assessments with respect to credibility, conflicting testimony, and the jury's choice of the competing inferences that can be drawn from the evidence, specious inferences are not indulged, because it would not satisfy the Constitution to have a jury determine that the defendant is probably guilty. If the evidence viewed in the light most favorable to the prosecution gives equal or nearly equal circumstantial support to a theory of guilt and a theory of innocence, then a reasonable jury must necessarily entertain a reasonable doubt.
To sustain a conspiracy conviction, the prosecution must prove beyond a reasonable doubt that the person charged with conspiracy knew of its existence and knowingly joined and participated in it. United States v. Rodriguez, 392 F.3d 539, 545 (2d Cir.2004). The Government must also prove, beyond a reasonable doubt, that the defendant possessed the specific intent to commit the offense that was the object of the conspiracy—here, kidnapping. United States v. Torres, 604 F.3d 58, 65 (2d Cir.2010). This requirement is contextual: the prosecution's proof must be considered in relation to the rest of the evidence presented at trial, rather than in isolation. Anderson, 747 F.3d at 59.
At trial, the prosecution built its case around Valle's chats and emails with his alleged co-conspirators. On appeal, it argues that these communications, “taken at face value, were fully sufficient to establish his intent to join a kidnapping conspiracy.” Gov't Opening Br. 32. We disagree.
As previously explained, Valle's chats and emails with the three alleged co-conspirators were part of a much larger set of chats and emails with 24 individuals on DFN. According to the prosecution, the former were unique because they evinced “real” criminal intent while the rest did not. After reviewing the chats and emails introduced at trial, the district court concluded that the “real” and “fantasy” chats were indistinguishable. 301 F.R.D. at 86.
Our review of the record yields the same conclusion. In both groups of chats, Valle transmits Facebook images of women and offers to kidnap and sell them on a “cash upon delivery” basis, and in both groups he expresses a desire to kidnap, rape, torture, and eat women whom he knows. In both groups Valle also claims to conduct surveillance of potential victims and discusses his intentions to kidnap them using chloroform and ropes. And in both groups he describes the various devices he “owns” that will assist in the process. Many of the “fantasy” chats also do not explicitly state that the participants are engaged in fantasy and are as graphic and detailed as the “real” chats. For example, the “real” chats and the “fantasy” chats both include haggling over the kidnapping fees that Valle “wanted to charge,” although the prosecution argues that this haggling is unique to the “real” conspiracy with VanHise. See id. at 84. The “real” chats thus contain the same core elements as the chats the Government concedes are “fantasy.”
In a “fantasy” chat with “Tim Chase,” for example, Valle and Chase agree to kidnap a woman on January 27, 2012. Valle supplies real pictures of the woman, they agree upon a price of $4,000 for Valle's services, and Valle states that the woman goes to the gym nightly and that he has kept a log of when she leaves and returns home. The two also agree upon a location “a hundred miles east of Erie” as the place of delivery. As Judge Gardephe found, there are no material differences between these chats with Chase, the fantasist, and Valle's chats with VanHise, the alleged co-conspirator. 301 F.R.D. at 86–87.
As noted earlier, Valle and Bollinger discussed the lunch over the course of multiple conversations and expressly referred to it as a way to confirm their choice of Sauer and inspire ideas for cannibalizing her. See supra at 532. In light of these discussions, it is not clear how the majority can contend that Valle and Bollinger's communications “leading up to and following the lunch” make it “impossible to conclude” that the lunch was in furtherance of their plot. Maj. Op. at 522.
Moreover, the “real” chats take place in the same time period as the admittedly “fantasy” chats. On the evening of July 12, 2012, for instance, Valle discusses kidnapping Andria Noble with Aly Khan in a “real” chat and, an hour later, discusses kidnapping Noble with someone else in a chat that was “fantasy.” The prosecution thus proposed that Valle simultaneously agreed to kidnap Noble while also engaging in role-play about the same woman. This temporal proximity casts further doubt upon any rational distinction between the chats.
Even when “taken at face value,” the “real” chats contain numerous other indicia of fantasy. For example, the prosecution alleged that Valle formed a genuine agreement with the specific intent to kidnap three different women in three different locations on the same day. First, Valle agreed with Aly Khan to lure Mangan to either India or Pakistan on February 20, 2012 and to slaughter her there. Second, he agreed with VanHise to kidnap Alisa Friscia in Manhattan on February 20, 2012 and deliver her to an unknown location in exchange for $4,000 in cash. Finally, Valle agreed with Aly Khan to kidnap Andria Noble on February 20, 2012 from her home in Columbus, Ohio. On appeal, the prosecution posits that the jury could have reasonably concluded that Valle seriously planned to kidnap Mangan, Friscia, and Noble on the same day and failed to go through with the kidnappings only because “an obstacle arose,” or because he had a “fear of getting caught.” Gov't Opening Br. 54. We believe that no rational juror could reach this conclusion for the reason noted by Judge Gardephe: “The notion that Valle had resolved to lure Mangan to India or Pakistan [to slaughter with Aly Khan,] while at the same time kidnapping Andria Noble in Columbus, Ohio, and kidnapping Alisa Friscia from the Upper East Side of Manhattan, is simply outlandish.” 301 F.R.D. at 90.
In addition to plots that would put the same person in different places at the same time, the “real” chats are replete with references to fantastical elements such as a human-sized oven, a spit, and a remote cabin in the woods, none of which Valle owned or made any effort to acquire. The fantastical nature of the “real” chats is bolstered by the entirely virtual nature of the alleged conspirators' relationships. Valle had no pre-existing relationship with those with whom he chatted, and he formed no real life relationship with any of them. He did not know their real names and, indeed, could not be sure of their genders, ages, or locations. Neither he nor his alleged co-conspirators made any effort to communicate by telephone, text message, or web camera, much less meet in person. And weeks or months could go by between Valle's chats with any particular individual. While anonymity is not uncommon in Internet communications, the fantastical elements of the chats combined with the impersonal nature of the interactions provides pervasive and unmistakable indicia of deep fantasy.
As Judge Richard Posner observed in another case involving an individual engaged in sexually graphic online communications with strangers, the defendant “may have thought (this is common in Internet relationships) that they were both enacting a fantasy.” United States v. Gladish, 536 F.3d 646, 650 (7th Cir.2008). Indeed, in Gladish the Seventh Circuit reversed a conviction for attempt to induce a minor to engage in sexual activity because “[the defendant's] talk and his sending her a video of himself masturbating (the basis of his unchallenged conviction for violating 18 U.S.C. § 1470)[we]re equally consistent with his having intended to obtain sexual satisfaction vicariously.” Id.
Unlike the majority, the jury did not determine that Valle was merely “interested in committing acts of sexualized violence against women,” Maj. Op. at 519; it concluded—beyond a reasonable doubt—that Valle actually conspired to commit a kidnapping.
Consequently, we need look no further than the prosecution's own work product to find reasonable doubt. The prosecution divided the exchanges into two groups and undertook to convince the jury to convict Valle on the theory that one group was fantasy and the other proved criminal intent. This exercise failed because the distinction the prosecution urged does not exist in this case. There is simply no material difference between the two groups of chats. We do not believe that the prosecution satisfies the proof beyond a reasonable doubt standard by relying upon a distinction that is untethered to reason or common sense.
Perhaps realizing that there is no actual distinction to be drawn between the “real” and “fantasy” chats, the prosecution now contends that it “did not take a position one way or the other as to whether [Valle's online communications with people other than the named co-conspirators] constituted genuine planning, puffery, preparatory conversations, role-playing, or something else entirely.” Gov't Opening Br. 39. The record, however, belies this assertion.
As noted, Agent Walsh was a key witness in the prosecution's case. He was the lead investigative agent and a majority of the chats and emails introduced were admitted into evidence through his testimony. He unequivocally testified, often in response to the Government's own questions, that the Government considered Valle's chats with 21 other individuals to be “fantasy” and Valle's chats with the three alleged co-conspirators to be “real.” The following exchanges between Agent Walsh and AUSA Hadassa Waxman on direct examination are illustrative:
WAXMAN: When you were reviewing those emails between [Valle] and the two dozen individuals, did you separate them into groups?
WALSH: I did.
WAXMAN: What were those groups?
WALSH: Ones that I believe that were real and ones that I believe were fantasy.
WAXMAN: Why did you make that separation?
WALSH: In the ones that I believe were fantasy, the individuals said they were fantasy. In the ones that I thought were real, people were sharing ... real details of women, names, what appeared to be photographs of the women, details of past crimes and they also said that they were for real.
WAXMAN: What caused you to make that separation between the emails you found that had realistic characteristics and those that were fantasy?
WALSH: Only my interest in obtaining information about that real criminal activity.
WAXMAN: Agent Walsh, approximately how many of Officer Valle's emails and electronic chats did you review in connection with your investigation?
WAXMAN: We just reviewed over yesterday and today about 40, is that right?
WALSH: That's correct, ma'am.
WAXMAN: Why did you focus on these 40 particular communications?
WALSH: We believed that these chats and e-mails contained elements of real crimes.
WAXMAN: And why did you come to that conclusion?
WALSH: They described dates, names, and activities that you would use to conduct a real crime.
WAXMAN: And did you cast aside a certain number of emails as well?
WALSH: Yes, ma'am.
WAXMAN: Why did you choose not to focus on those emails?
WALSH: Quite frankly, ma'am, they didn't seem realistic.
WAXMAN: Why not?
WALSH: They were clearly role-play. They used the word “fantasy” in the actual chats or emails.
SA 125–26. On cross-examination, Agent Walsh admitted that the grouping decision was made by numerous agents and prosecutors.
BAUM: Now, when you made that decision that 21 out of 24 participants with Mr. Valle were engaged in fantasy role-play, were you the only one who made that decision?
WALSH: No, sir.
BAUM: How many agents were involved in that decision?
WALSH: Approximately eight to 10, sir.
BAUM: And how many people from the U.S. Attorney's Office were involved in that decision?
WALSH: About two, sir.
BAUM: So eight to 10 law enforcement officers and at least two lawyers from the U.S. Attorney's Office decided that out of 24 people that Mr. Valle chatted or emailed with[,] 21 were fantasy role-plays, is that correct?
WALSH: Approximately. Yes, sir.
SA 129–30. The prosecution now urges that the distinction between “real” chats and “fantasy” role play was Valle's defense and that the district court applied the wrong standard by forcing the prosecution to disprove the defense theory of the case. As the exchanges above demonstrate, the distinction was introduced and relied on by the Government's case agent. In any event, intent is an essential element of the crime that the Government charged. The issue, therefore, is not whether the prosecution disproved the defense 's theory, but whether the prosecution proved its theory that Valle's “real” chats represented a departure from his otherwise entirely imaginary world.
Alternatively, the Government argues that even if it introduced the distinction, it did not rely on or concede the truth of the distinction because it “did not even introduce any of the ‘fantasy’ conversations at trial so that [a] comparison could be made” with the “real” chats. Gov't Opening Br. 41. A sampling of the “fantasy” chats was introduced by the defense in its cross-examination of Agent Walsh. But regardless of how the exhibits were introduced, the Government's own investigation concluded that forty chats permitted the inference of conspiratorial intent, as compared to myriad other chats that did not. The Government claims that it does not have to prove a distinction between these two sets of chats because the jury could have rationally found that “defendants charged with attempting or conspiring to engage in criminal, deviant activity often contemporaneously engage in ‘fantasy’ behavior ... about activity ... that is similar to the charged conduct.” Id. at 43.
This contention proves too little. Once the Government constructs its case around the theory that a certain group of chats permits the inference of conspiratorial intent while another group of essentially similar chats is consistent with non-criminal behavior, some adequate explanation must be forthcoming. Where, as here, none is, the non-criminal chats are a powerful indicator that a reasonable juror must necessarily entertain reasonable doubt about the prosecution's case.
Unable to materially distinguish the “real” chats from the “fantasy” chats, the Government relies on evidence of “real world” steps that Valle took in order to “prepare” for the kidnappings. See, e.g., Gov't Opening Br. 56–58, 60–68. For example, the prosecution introduced evidence that Valle performed Internet searches for how to kidnap people, how to make chloroform, and how to restrain and cannibalize people. The prosecution also introduced evidence that Valle researched prior kidnappings, which it argues permitted the jury to infer that Valle was interested in how those kidnappers were caught so that he could learn from their experiences and avoid apprehension. Id. at 63.
To be sure, Internet searches can provide some relevant proof of intent. However, an Internet search, in and of itself, is not criminal. Here, the searches on which the Government relies occurred in a context of deep fantasy. As with his chats and emails, Valle's Internet searches show that he was interested in committing acts of sexualized violence against women. Interest may be relevant evidence of intent, but it does not by itself prove intent. “No doubt some people commit sex crimes because they want to turn their fantasies into reality, but most people with criminal fantasies probably refrain from acting on them, because they know it would be wrong, or because they do not want to risk the penalties.” Curtin, 489 F.3d at 962 (Kleinfeld, J., concurring).
The Government also relies on at least two occasions when Valle engaged in acts of “surveillance” of his intended victims. First, the Government notes that Valle admitted in a post-arrest statement that he was on Friscia's block on March 1, 2012, two days after he allegedly agreed to kidnap her with VanHise. Gov't Opening Br. 56–58. Valle told a government agent that he was on the block to drop off Mangan to have lunch with Friscia, but both Mangan and Friscia testified that they had not met for lunch that day. Valle indicated to the agent that he was on the block only very briefly, and there is no evidence to the contrary. There is also no evidence that he observed Friscia or her apartment building while he was on her block. Valle's false exculpatory explanation for being on the block is “insufficient proof on which to convict where other evidence of guilt is weak.” United States v. Johnson, 513 F.2d 819, 824 (2d Cir.1975). As the district court found, no rational juror could conclude from this evidence alone that Valle was engaged in “surveillance.”
Second, the prosecution and our dissenting colleague contend that the jury could convict Valle of a conspiracy to kidnap based on his communications with Moody Blues about Kimberly Sauer. See Gov't Opening Br. 35–36. This evidence is insufficient to show that Valle agreed or had the specific intent to kidnap Sauer and, in any event, it does not establish Moody Blues's intent.
Sauer is a former college classmate of Valle's who lives in Maryland. According to Sauer, she communicated with Valle by text message approximately ten to fifteen times a year. Mangan testified that she and Valle made three or four trips to Maryland during the course of their relationship (from 2009 through September 2012) and that each time she and Valle made an effort to see Sauer when in the area.
In January 2012, Valle asked Sauer for her address so that he could send her a Patrolmen's Benevolent Association card. The earliest chat between Valle and Moody Blues introduced at trial takes place seven months later, on July 9, 2012. During this conversation, Valle described several girls that he was “working on grabbing ... for thanksgiving,” and told Moody Blues that “Kimberly [is] by far the easiest” to kidnap because he could “just show up at her home unannounced.” JA 80–82. After Valle suggested that “maybe you can make it here and help me with her, since you have experience,” Moody Blues responded that he lives in England but it is “easy to get to the Big apple.” JA 81.
Valle also told Moody Blues that he was “single,” had a “big gas oven,” and that “no one is around [him] for about ¾ of a mile.” JA 84. The two then discussed how they would truss up Sauer and cook her on an outdoor spit at Valle's mountain house. During this same chat, approximately one hour after Valle wrote that he wanted to kidnap someone for Thanksgiving, Valle told Moody Blues that he was “thinking of a Labor Day cookout ... with Kimberly as the main course.” JA 86. Valle noted that she had “been one of my favorite victims to fantasize about for almost 10 years now.” JA 86. Again during the same chat in early July, Valle sent Moody Blues a link to a video of Sauer on vacation and volunteered to make chloroform and buy rope. Moody Blues replied that “Labour day is the 3rd [of] September, not a lot of time to sort out plane tickets etc. Will see what cheap deals I can get.” JA 90.
One day later, on July 10, Valle sent Moody Blues “a word document, a blueprint of everything we will need to carry this out.” JA 100. The document, entitled “Abducting and Cooking Kimberly: A Blueprint,” has a “target date” of September 2, 2012 for the abduction. It includes a photograph of Sauer, and accurately describes her age and marital status and that she is not a drug user, does not have tattoos, and drinks only occasionally. All of the other information in this document is false, including her last name, date of birth, birthplace, and educational history. The entire “plan” for abduction set out in the “Blueprint” is as follows: “I will arrive at some point Sunday night at her home to kidnap her. She lives in a quiet suburban neighborhood (Pictures of her house to be added).” The document also lists some materials that are needed, including a car, chloroform, rope, gag, tarp/plastic bags, gloves, and cheap sneakers. JA 267–68. After receiving the “Blueprint,” Moody Blues asked “[m]ay I have her address? For Googling using the Map app?” JA 101. Valle lied that he was “not sure” of her exact address. Id. There is no evidence in the record that Valle ever obtained any of the materials listed in the “Blueprint,” or that the document was ever updated with pictures of Sauer's house or any additional information.
At some point prior to July 12, Valle called Sauer to tell her that he would be traveling to Maryland with his wife and daughter for a weekend. They made plans to meet for lunch on July 22. On July 17, Valle informed Moody Blues that he would be having lunch with Sauer. Later in this chat, Moody Blues asked Valle if he had a recipe for chloroform. Valle sent him a link. On July 19, Moody Blues again asked for Sauer's address, and Valle replied that he did not know it by heart. Valle never provided Moody Blues with Sauer's address.
On July 20, Valle conducted a number of Internet searches relating to kidnapping, including “how to kidnap someone,” “how to chloroform a girl,” and “kidnapped girl.” On July 21, Valle traveled to Maryland with his wife and daughter. They visited several college friends, and had the scheduled lunch with Sauer on July 22. On July 21, Valle texted Sauer “[w]e drove by your pink building today,” and she responded “Haha yay!” JA 237. At trial, Sauer testified that she understood Valle to refer to her office building, which has pink-tinted windows, but that Valle had never visited her at work and she had never sent him photographs of the building. She described the lunch as “fine” and “pleasant.”
On the evening of July 22, after Valle returned home, he emailed Moody Blues that Sauer “looked absolutely mouthwatering.” JA 117. Valle and Moody Blues said nothing more about the plot to kidnap Sauer and did not talk again for another month. On August 21, Valle and Moody Blues began to discuss Kristen Ponticelli, a recent graduate of Valle's high school whom he did not know. JA 264. There is no evidence in the record that Valle and Moody Blues ever discussed Sauer or Ponticelli again after August 21.
As Judge Gardephe observed, the chats pertaining to Sauer are not materially different from the other fantasy chats. All of the elements of this alleged plot are equally fantastical, including the presence of the nonexistent mountain house, the human-sized oven, and the “Blueprint.” The “plan” to kidnap Sauer in the “Blueprint” is no more detailed than is the “plan” in Valle's Internet chats with Moody Blues, nor does the list of materials required differ from the types of materials Valle discusses in his chats. And critically, Valle makes concerted efforts to conceal from Moody Blues any identifying information about Sauer that could be used in furtherance of a kidnapping such as her last name, date of birth, and the name of her alma mater. Although the prosecution speculates that Valle did not share accurate information about Sauer because he did not want Moody Blues to undertake the kidnapping without him, there is no evidence in the record to support such an inference.
Thus, the only meaningful difference between this alleged conspiracy and the “fantasy” chats is the occurrence of Valle's lunch with Sauer in Maryland during approximately the same time period as he discussed kidnapping her with Moody Blues. Although the Government characterizes Valle's communications with Sauer as “out of the blue,” the record shows that they communicated by text message in the year prior to the alleged kidnapping plot on a regular basis and that they made an effort to see each other when Valle was in town. Valle did not have lunch with Sauer alone, but rather came with Mangan and their infant daughter. Moreover, the chats between Moody Blues and Valle leading up to and following the lunch make it impossible to conclude, without speculation, that the lunch was “surveillance” in furtherance of a genuine conspiracy. Moody Blues makes only a single reference to purchasing plane tickets in the July 9 chat, but that suggestion is never brought up again. Except for the e-mail recapping the lunch on July 22 and the August 24 conversation in which their focus moves to Ponticelli after a brief mention of Sauer, Moody Blues and Valle never again discuss Sauer or any plot to kidnap her. In fact, Moody Blues and Valle do not speak at all for the month after July 22, and the “target date” of September 2 passes with no discussion. And Valle never takes any step of any sort in furtherance of an alleged kidnapping.
We are in accord with the prosecution and our dissenting colleague that a jury might be able to distill some incriminating evidence from all of this. But “some” evidence is not the test. Because Valle's relationship with Moody Blues is essentially indistinguishable from his relationship with all of the others with whom he chatted, we agree with Judge Gardephe that a rational jury could not conclude that this evidence was sufficient to meet the “beyond any reasonable doubt” requirement. As our case law instructs:
[I]t is not enough that the inferences in the government's favor are permissible. A court must also be satisfied that the inferences are sufficiently supported to permit a rational juror to find that [each element of the offense] is established beyond a reasonable doubt.
If the evidence viewed in the light most favorable to the prosecution gives equal or nearly equal circumstantial support to a theory of guilt and a theory of innocence, then a reasonable jury must necessarily entertain a reasonable doubt.
United States v. Triumph Capital Grp., Inc., 544 F.3d 149, 159 (2d Cir.2008).
Finally, on the basis of this evidence, it is impossible to determine beyond a reasonable doubt whether Moody Blues—or for that matter any of Valle's other alleged co-conspirators—ever had the specific intent to commit a kidnapping. We have taken a bilateral approach to the crime of conspiracy: at least two people must agree. “When one of two persons merely pretends to agree, the other party, whatever he may believe, is in fact not conspiring with anyone.” See United States v. Bicaksiz, 194 F.3d 390, 398 (2d Cir.1999). The only evidence the Government offers to demonstrate Moody Blues's intent is the words he used in the chats. Gov't Reply Br. 21–22. As we have explained, these chats of “real” criminal intent are rife with indicia of fantasy and contain the same substantive elements as the chats the Government concedes are “fantasy.” The conclusion that the chats do not support a finding of Valle's conspiratorial intent applies with equal force to Moody Blues.
On this record, no reasonable juror could conclude beyond a reasonable doubt that Valle possessed the specific intent to kidnap anyone or that he and his alleged co-conspirators ever formed an agreement to actually carry out any of the purported kidnappings. The mere indulgence of fantasy, even of the repugnant and unsettling kind here, is not, without more, criminal. We therefore affirm the district court's judgment of acquittal as to the conspiracy count.
We now turn to Valle's appeal of the judgment of conviction on the CFAA count. We reverse because section 1030(a)(2)(B) is ambiguous and where, as here, the Government and the defense both posit plausible interpretations of a criminal statute, the rule of lenity requires us to adopt the defendant's construction. As Justice Scalia has emphasized, “[w]hen interpreting a criminal statute, we do not play the part of a mindreader.” United States v. Santos, 553 U.S. 507, 515, 128 S.Ct. 2020, 170 L.Ed.2d 912 (2008). When “a reasonable doubt persists about a statute's intended scope even after resort to the language and structure, legislative history, and motivating policies of the statute,” Moskal v. United States, 498 U.S. 103, 108, 111 S.Ct. 461, 112 L.Ed.2d 449 (1990), we resolve doubts in favor of the defendant rather than “imputing to Congress an undeclared will” to criminalize conduct, Santos, 553 U.S. at 515, 128 S.Ct. 2020 (quoting Bell v. United States, 349 U.S. 81, 83, 75 S.Ct. 620, 99 L.Ed. 905 (1955)). The rule of lenity ensures that criminal statutes will provide fair warning of what constitutes criminal conduct, minimizes the risk of selective or arbitrary enforcement, and strikes the appropriate balance between the legislature and the court in defining criminal liability. See Yates v. United States, –––U.S. ––––, 135 S.Ct. 1074, 1088, 191 L.Ed.2d 64 (2015); United States v. Simpson, 319 F.3d 81, 86 (2d Cir.2002).
The CFAA imposes criminal and civil liability on one who, among other things, “intentionally accesses a computer without authorization or exceeds authorized access and thereby obtains information ... from any department or agency of the United States.” 18 U.S.C. § 1030(a)(2)(B). “Without authorization” is not defined. However, “ ‘exceeds authorized access' means to access a computer with authorization and to use such access to obtain or alter information in the computer that the accesser is not entitled so to obtain or alter.” Id. § 1030(e)(6).
The dispositive question is whether Valle “exceeded authorized access” when he used his access to OFM to conduct a search for Maureen Hartigan with no law enforcement purpose. Valle concedes that he violated the terms of his employment by putting his authorized computer access to personal use, but claims that he did not violate the statute because he never “used his access to obtain any information he was not entitled to obtain.” Valle's Opening Br. 8. In other words, Valle argues that he did not “exceed authorized access” because he was otherwise authorized to obtain the database information about Hartigan; his non-law enforcement purpose in running the search is irrelevant. See id. at 9. The Government contends that Valle “exceeded authorized access” because his authorization to access OFM was limited to law enforcement purposes and he conducted a search for Hartigan with no such purpose.
The critical term—“authorization”—is not defined in the statute, but we have previously recognized in construing the CFAA that “authorization” is a word “of common usage, without any technical or ambiguous meaning.” United States v. Morris, 928 F.2d 504, 511 (2d Cir.1991). The dictionary defines “authorization” as “permission or power granted by authority.” Random House Unabridged Dictionary 139 (2001). Thus, common usage of “authorization” suggests that one “accesses a computer without authorization” if he accesses a computer without permission to do so at all. See, e.g., LVRC Holdings LLC v. Brekka, 581 F.3d 1127, 1133 (9th Cir.2009).
See also Black's Law Dictionary 159 (10th ed.2014) (defining “authorization” as “[o]fficial permission to do something”); Webster's Third International Dictionary 146 (2002) (defining “authorization” as “the state of being authorized,” and “authorize” as “to endorse, empower, justify, permit by or as if by some recognized or proper authority”).
This evidence, viewed in the light most favorable to the government and “in its totality,” United States v. Florez, 447 F.3d 145, 154 (2d Cir.), cert. denied, 549 U.S. 1040, 127 S.Ct. 600, 166 L.Ed.2d 445 (2006), is not merely “some incriminating evidence,” Maj. Op. at 522, whatever that may mean.
Common usage of “authorization” is less helpful in determining when one “exceeds authorized access” because it can support both Valle's and the Government's interpretation. While “authorization” could refer, as the Government contends, to the purposes for which one is authorized to access a computer, it could alternatively refer to the particular files or databases in the computer to which one's authorization extends. Indeed, by its plain terms the statute is directed to improper “access.” The contested language is not “exceeds authorization,” however such authorization may be defined, but the seemingly more limited “exceeds authorized access.” 18 U.S.C. § 1030(a)(2) (emphasis added). Moreover, because “without authorization” most naturally refers to a scenario where a user lacks permission to access the computer at all, one sensible reading of the statute is that “exceeds authorized access” is complementary, referring to a scenario where a user has permission to access the computer but proceeds to “exceed” the parameters of authorized access by entering an area of the computer to which his authorization does not extend. As Judge Kozinski recognized in United States v. Nosal, “it is possible to read both prohibitions as applying to hackers: ‘Without authorization’ would apply to outside hackers (individuals who have no authorized access to the computer at all) and ‘exceeds authorized access' would apply to inside hackers (individuals whose initial access to a computer is authorized but who access unauthorized information or files).” 676 F.3d 854, 858 (9th Cir.2012) (en banc).
Over the past fourteen years, six other circuits have wrestled with the question before us. Most recently, the Ninth Circuit sitting en banc in Nosal and the Fourth Circuit in WEC Carolina Energy Solutions LLC v. Miller, 687 F.3d 199 (4th Cir.2012), adopted Valle's construction. Before that, the First, Fifth, Seventh, and Eleventh Circuits adopted the prosecution's interpretation. See United States v. John, 597 F.3d 263 (5th Cir.2010); United States v. Rodriguez, 628 F.3d 1258 (11th Cir.2010); Int'l Airport Ctrs., L.L.C. v. Citrin, 440 F.3d 418 (7th Cir.2006); EF Cultural Travel BV v. Explorica, Inc., 274 F.3d 577 (1st Cir.2001). If this sharp division means anything, it is that the statute is readily susceptible to different interpretations. We therefore turn to the legislative history and motivating policies for further guidance.
The dissent claims that we “discover[ ] ambiguity in the statutory language where there is none” and summarily concludes that “exceeds authorized access” obviously encompasses a scenario where a user “did not comply with restrictions on [his] authorized access.” Dissenting Op. at 537–38. This conclusion is, with respect, not reasonable in light of these cases.
These concerns address Section 1030(a)(2)(C) of the CFAA, which prohibits exceeding authorized access to a computer to obtain “information from any protected computer.” 18 U.S.C. § 1030(a)(2)(C). The CFAA defines “protected computer” as, among other things, a computer “which is used in or affecting interstate or foreign commerce or communication,” id. § 1030(e)(2)(B), a definition that has been interpreted to encompass any computer with an internet connection, see United States v. Yücel, 97 F.Supp.3d 413, 418–19 (S.D.N.Y.2015) (collecting cases and noting “widespread agreement in the case law” that “protected computer” includes any internet-connected computer). Valle was not charged under Section 1030(a)(2)(C).
Congress enacted the CFAA in 1984 to address “computer crime,” which was then principally understood as “hacking” or trespassing into computer systems or data. See H.R.Rep. No. 98–894, at 3691–92, 3695–97 (1984); S.Rep. No. 99–432, at 2480 (1986). The House Committee Report to the original bill detailed the existence of “ ‘hackers' who have been able to access (trespass into) both private and public computer systems” as a result of the “corresponding proliferation of computer networking which began during the 1970's.” H.R.Rep. No. 98–894, at 3695; see also id. at 3696 (noting the “recent flurry of electronic trespassing incidents”). The report described one instance of “computer crime” in which an individual “stole confidential software by tapping into the computer system of a previous employer from [the] defendant's remote terminal.” Id. at 3691–92.
The Senate Committee Report to the 1986 amendments specifically described “exceeds authorized access” in terms of trespassing into computer systems or files. In heightening the mens rea requirement for section 1030(a)(2), the Committee explained that it did not want to hold liable those “who inadvertently ‘stumble into’ someone else's computer file or computer data,” which was “particularly true in those cases where an individual is authorized to sign onto and use a particular computer, but subsequently exceeds his authorized access by mistakenly entering another computer or data file that happens to be accessible from the same terminal.” S.Rep. No. 99–432, at 2483. Congress was also careful to note that “section 1030 deals with an ‘unauthorized access' concept of computer fraud rather than the mere use of a computer. Thus, the conduct prohibited is analogous to that of ‘breaking and entering.’ ” H.R.Rep. No. 98–894, at 3706. Consequently, the legislative history consistently characterizes the evil to be remedied—computer crime—as “trespass” into computer systems or data, and correspondingly describes “authorization” in terms of the portion of the computer's data to which one's access rights extend.
The Government relies upon the predecessor language to “exceeds authorized access.” As originally enacted, section 1030(a) made it a crime to “knowingly access[ ] a computer without authorization, or having accessed a computer with authorization, use[ ] the opportunity such access provides for purposes to which such authorization does not extend.” Counterfeit Access Device and Computer Fraud and Abuse Act of 1984, Pub.L. No. 98–473, § 2102(a), 98 Stat. 1837, 2190 (codified as amended at 18 U.S.C. § 1030) (emphasis added). In 1986, Congress deleted the italicized language and replaced it with “exceeds authorized access.” S.Rep. No. 99–432, at 2486.
The Government argues that no substantive change was intended because the substitution was made “to simplify the language.” Id. Valle cites another provision of the Senate Committee Report, relating to subsection (a)(3), which states that Congress had “eliminate[d] coverage for authorized access that aims at ‘purposes to which such authorization does not extend,’ ” and thereby “remove[d] from the sweep of the statute one of the murkier grounds of liability under which a[n] ... employee's access to computerized data might be legitimate in some circumstances, but criminal in other (not clearly distinguishable) circumstances.” Id. at 2494 (emphasis added). He argues that Congress therefore intended to abrogate any purpose-based inquiry by substituting the new “exceeds authorized access” language. While a number of courts have found this argument persuasive, see, e.g., Nosal, 676 F.3d at 858 n. 5, we have misgivings. It seems more likely that the Committee was merely explaining its removal of “exceeds authorized access” as a basis for liability under subsection (a)(3), rather than the substitution of “exceeds authorized access” in other provisions of the statute, including subsection (a)(2).
Nevertheless, we do not think that the appearance of the word “purposes” in the legislative history renders the statute clear for the simple reason that even when Congress referenced the user's “purposes,” it spoke in terms of the particular computer files or data to which the user's access rights extended. The Committee's extensive discussion of subsection (a)(3) is instructive. As initially enacted, that provision made it a crime to knowingly access a government computer without authorization or exceed the scope of one's authorization and thereby use or disclose information. S.Rep. No. 99–432, at 2494. Subsection (a)(3) therefore “swe[pt] in all computerized government information, including documents that must, under the Freedom of Information Act [ (“FOIA”) ], be disclosed to any member of the public upon proper request,” while “gloss[ing] over the reality that the existence or exact scope of a government employee's authority to access a particular computerized data base is not always free from doubt.” Concerned that government employees would “resolve doubts against disclosure” when responding to FOIA requests, the Committee revised subsection (a)(3) in three ways, including by removing the “purposes” language. Id.
Each of these revisions was directed toward the same problem: an employee with authorization to access certain databases entering other databases to which his authorization did not extend. And, in explaining the revisions, the Committee understood authorization in spatial terms, namely, an employee going beyond the parameters of his access rights. See e.g., id. at 2495 (declining to apply subsection (a)(3) “to access by a Federal employee of computers of that employee's own agency,” and explaining that the revised rule “would provide prosecutors a clear, workable rule, regardless of the intricacies of a particular agency's computer access policies: absent a fraudulent motive, an employee could not be prosecuted for simple ‘trespass ’ into one of his agency's own computers”) (emphasis added). This understanding of authorization is, as we have previously explained, consistent with Congress's discussion of the concept elsewhere. It is likewise consistent with the statute's principal purpose of addressing the problem of hacking, i.e., trespass into computer systems or data.
At the end of the day, we find support in the legislative history for both Valle's and the Government's construction of the statute. But because our review involves a criminal statute, some support is not enough. Where, as here, ordinary tools of legislative construction fail to establish that the Government's position is unambiguously correct, we are required by the rule of lenity to adopt the interpretation that favors the defendant. Santos, 553 U.S. at 514, 128 S.Ct. 2020; United States v. Granderson, 511 U.S. 39, 54, 114 S.Ct. 1259, 127 L.Ed.2d 611 (1994). We do not think it too much to ask that Congress define criminal conduct with precision and clarity. As Chief Justice Marshall explained:
The rule that penal laws are to be construed strictly, is perhaps not much less old than construction itself. It is founded ... on the plain principle that the power of punishment is vested in the legislative, not in the judicial department. It is the legislature, not the Court, which is to define a crime, and ordain its punishment.
United States v. Wiltberger, 18 U.S. (5 Wheat.) 76, 95, 5 L.Ed. 37 (1820). We decline to adopt the prosecution's construction, which would criminalize the conduct of millions of ordinary computer users and place us in the position of a legislature.
The role that the rule of lenity plays where doubt remains as to the reach of a criminal statute was discussed in Nosal, where the Ninth Circuit sitting en banc focused sharply on the same compelling concerns that Valle and amici raise on this appeal:
[T]he government's proposed interpretation of the CFAA allows private parties to manipulate their computer-use and personnel policies so as to turn these relationships into ones policed by the criminal law. Significant notice problems arise if we allow criminal liability to turn on the vagaries of private policies that are lengthy, opaque, subject to change and seldom read. Consider the typical corporate policy that computers can be used only for business purposes. What exactly is a ‘nonbusiness purpose’? If you use the computer to check the weather report for a business trip? For the company softball game? For your vacation to Hawaii? And if minor personal uses are tolerated, how can an employee be on notice of what constitutes a violation sufficient to trigger criminal liability?
676 F.3d at 860. The Fourth Circuit, in Miller, agreed with the Ninth Circuit and echoed the same concerns:
The deficiency of a rule that revokes authorization when an employee uses his access for a purpose contrary to the employer's interests is apparent: Such a rule would mean that any employee who checked the latest Facebook posting or sporting event scores in contravention of his employer's use policy would be subject to the instantaneous cessation of his agency and, as a result, would be left without any authorization to access his employer's computer systems.... [W]e do not think Congress intended ... the imposition of criminal penalties for such a frolic.
We agree with the Ninth and Fourth Circuits that courts that have adopted the broader construction “looked only at the culpable behavior of the defendants before them, and failed to consider the effect on millions of ordinary citizens caused by the statute's unitary definition of ‘exceeds authorized access.’ ” Nosal, 676 F.3d at 863; see also Miller, 687 F.3d at 206 (“[W]e believe that th[is] theory has far-reaching effects unintended by Congress.”). This is the very concern at the heart of the rule of lenity.
For example, in United States v. Kozminski, 487 U.S. 931, 108 S.Ct. 2751, 101 L.Ed.2d 788 (1988), the Supreme Court refused to adopt the Government's broad interpretation of a statute criminalizing involuntary servitude. The Government argued that the statute should criminalize “compulsion [to work] through psychological coercion as well as almost any other type of speech or conduct intentionally employed to persuade a reluctant person to work.” Id. at 949, 108 S.Ct. 2751. The Supreme Court rejected this interpretation because it would “criminalize a broad range of day-to-day activity,” such as “a parent who coerced an adult son or daughter into working in the family business by threatening withdrawal of affection.” Id. The Court warned that the broader statutory interpretation would “delegate to prosecutors and juries the inherently legislative task of determining what type of ... activities are so morally reprehensible that they should be punished as crimes” and would “subject individuals to the risk of arbitrary or discriminatory prosecution and conviction.” Id.
The Government does not reply substantively to Valle's concerns about the rule of lenity or about the risk of criminalizing ordinary behavior inherent in its broad construction. It merely states that “those concerns must be raised in the first instance by individuals actually affected by the provision at issue,” and that “[t]hose cases will present fact-specific questions not relevant here, including whether the applicable authorization was clearly defined and whether the abuse of computer access was intentional.” Gov't Opp'n Br. 15. We disagree. The Government asks that we affirm Valle's conviction, which requires us to accept its construction of the statute. But our construction of the statute impacts many more people than Valle. It will not only affect those who improperly access information from a government computer—a result some readers might find palatable—but also those who improperly access “any protected computer” and thereby obtain information. 18 U.S.C. § 1030(a)(2)(C). As the Ninth Circuit aptly put it in Nosal, “[b]ecause ‘protected computer’ is defined as a computer affected by or involved in interstate commerce—effectively all computers with Internet access—the government's interpretation of ‘exceeds authorized access' makes every violation of a private computer use policy a federal crime.” 676 F.3d at 859 (citing 18 U.S.C. § 1030(e)(2)(B)).
Whatever the apparent merits of imposing criminal liability may seem to be in this case, we must construe the statute knowing that our interpretation of “exceeds authorized access” will govern many other situations. See 18 U.S.C. § 1030(e)(6). It is precisely for this reason that the rule of lenity requires that Congress, not the courts or the prosecutors, must decide whether conduct is criminal. We, on the other hand, are obligated to “construe criminal statutes narrowly so that Congress will not unintentionally turn ordinary citizens into criminals.” Nosal, 676 F.3d at 863. While the Government might promise that it would not prosecute an individual for checking Facebook at work, we are not at liberty to take prosecutors at their word in such matters. A court should not uphold a highly problematic interpretation of a statute merely because the Government promises to use it responsibly. See United States v. Stevens, 559 U.S. 460, 480, 130 S.Ct. 1577, 176 L.Ed.2d 435 (2010).
For these reasons, we AFFIRM the judgment of acquittal as to the count of conspiracy to kidnap, and REVERSE the judgment of conviction as to the count of improperly accessing a computer in violation of the CFAA.
Judge STRAUB dissents in a separate opinion.
STRAUB, Circuit Judge, dissenting:
This case is important. It is important to the role we have traditionally allotted to juries in our criminal justice system. It is important to the rule which requires judges to apply the laws written by the people's Congress, particularly to police conduct which utilizes official databases to access information about citizens where such is not part of official action. Because the majority opinion seeks to enshrine all the conduct in this case in an academic protective halo, I find it necessary to offer the realistic context of this controversy.
This is not a case about governmental intrusion on one's personal inclinations and fantasies nor is it a case about governmental punishment of one's thoughts. It is, instead, a jury's determination of guilt for a conspiracy based on definitive conduct. This is not a case of confused, accidental, or otherwise inappropriate use of a law enforcement database. It is, instead, a police officer's use of the official database to obtain, outside the boundaries of his official duties, data about a woman whom he knew.
Having so noted the context of this appeal, I now proceed to presentation of the facts and applicable law.
* * *
Defendant–Appellant/Defendant–Appellee Gilberto Valle was convicted by a jury of conspiracy to kidnap (Count I) and of improperly accessing a computer in violation of the Computer Fraud and Abuse Act (“CFAA”) (Count II). Valle moved for a judgment of acquittal on both counts. As to Count I, the District Court (Paul G. Gardephe, Judge) granted Valle's motion, finding that the evidence at trial was insufficient to support the jury's verdict, and entered a judgment of acquittal. As to Count II, the District Court denied Valle's motion, rejecting Valle's argument that the CFAA did not proscribe his conduct, and entered a judgment of conviction.
The majority affirms the judgment of acquittal in respect of Count I and reverses the judgment of conviction in respect of Count II. I would instead vacate the District Court's judgment of acquittal in respect of Count I, affirm the District Court's judgment of conviction in respect of Count II, and remand for further proceedings. Therefore, I respectfully dissent.
I. Conspiracy to Kidnap (Count I)
The Sixth Amendment right to trial by jury is a “fundamental reservation of power in our constitutional structure,” Blakely v. Washington, 542 U.S. 296, 306, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004), and reflects our “deep commitment ... to the right of jury trial in serious criminal cases as a defense against arbitrary law enforcement,” Duncan v. Louisiana, 391 U.S. 145, 156, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968). “Just as suffrage ensures the people's ultimate control in the legislative and executive branches, jury trial is meant to ensure their control in the judiciary.” Blakely, 542 U.S. at 306, 124 S.Ct. 2531; see also United States v. Bailey, 444 U.S. 394, 435, 100 S.Ct. 624, 62 L.Ed.2d 575 (1980) (Blackmun, J., dissenting) (“The jury is the conscience of society and its role in a criminal prosecution is particularly important.”). In this “strict division of authority between judge and jury,” Blakely, 542 U.S. at 313, 124 S.Ct. 2531, it is the sole responsibility of the jury to weigh evidence and reach the ultimate conclusion of guilt or innocence, see United States v. Gaudin, 515 U.S. 506, 514, 115 S.Ct. 2310, 132 L.Ed.2d 444 (1995); Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979).
To preserve the jury's broad fact-finding discretion, a court can enter a judgment of acquittal only if the evidence of guilt is “nonexistent or so meager that no reasonable jury could find guilt beyond a reasonable doubt.” United States v. Espaillet, 380 F.3d 713, 718 (2d Cir.2004) (internal quotation marks omitted). In making this determination, a court must defer to the jury's resolution of evidentiary conflicts. Jackson, 443 U.S. at 319, 99 S.Ct. 2781; see also United States v. Hamilton, 334 F.3d 170, 179 (2d Cir.) (“[W]e must defer to the jury's resolution of the weight of the evidence and the credibility of the witnesses, and to the jury's choice of the competing inferences that can be drawn from the evidence.” (internal quotation marks, citations, and brackets omitted)), cert. denied, 540 U.S. 985, 124 S.Ct. 502, 157 L.Ed.2d 378 (2003). We therefore view the evidence in the light most favorable to the government and draw all reasonable inferences in its favor. Jackson, 443 U.S. at 319, 99 S.Ct. 2781; see also United States v. Kozeny, 667 F.3d 122, 139 (2d Cir.2011) (noting the “heavy burden” that a defendant faces when seeking to overturn a jury's verdict), cert. denied, ––– U.S. ––––, 133 S.Ct. 1794, 185 L.Ed.2d 810 (2013).
The majority blatantly disregards these constraints on a court's review of a jury's verdict and accords to itself the power to “identify” and “discern [ ]” “[t]he line between fantasy and criminal intent.” Maj. Op. at 511. Rather, it was, and remains, for the jury to determine the factual question of whether Valle had criminal intent. The jury considered and rejected Valle's defense that he was simply pretending to commit a crime. It instead found, beyond a reasonable doubt, that Valle actually and genuinely conspired to kidnap someone. The majority's eloquent prose on the importance of protecting thoughts from criminal punishment, see id. at 511—perhaps better suited for an opinion editorial than a judicial opinion—is thus irrelevant, because the jury did not convict Valle for fantasizing. We are left to determine only whether the evidence, viewed in the light most favorable to the government, was sufficient for the jury to reach its verdict.
In reaching its conclusion, the majority pays lip service to the standard we apply in evaluating the sufficiency of the evidence, but then usurps the jury's role by weighing competing inferences and explanations of the evidence rather than viewing it in the government's favor. Such an undertaking is plainly inappropriate in this context.
Perhaps most tellingly, the majority imprudently structures its analysis to support its predetermined outcome by viewing the evidence “in its parts,” rather than considering “the government's case in its totality,” United States v. Hawkins, 547 F.3d 66, 70 (2d Cir.2008) (internal quotation marks and brackets omitted): it first concludes that Valle's online communications—viewed in isolation—are “indistinguishable” from the purported “fantasy” communications, see Maj. Op. at 516, and then proceeds to reject the remaining evidence (of Valle's overt acts in furtherance of the conspiracy) as insufficient on its own to support the conviction, see id. at 521– 23. When considered properly with the deference that we must apply, however, the jury's guilty verdict on the count of conspiracy to kidnap was sufficiently supported by the evidence presented at trial.
A. The Evidence Supporting the Jury's Verdict, Viewed in the Light Most Favorable to the Government
To decide that Valle was guilty of conspiracy to kidnap, all that the jury had to find was sufficient evidence in respect of one alleged coconspirator and one intended victim. See Kozeny, 667 F.3d at 131–32; United States v. Thomas, 54 F.3d 73, 81 (2d Cir.1995). The evidence at trial provided a sufficient basis for the jury to conclude that Valle made an agreement with Dale Bollinger, a man he met on darkfetishnet.com, to kidnap and murder Kimberly Sauer, one of Valle's friends from college.
During the earliest communication in evidence between Valle and Bollinger, Valle sought Bollinger's assistance kidnapping and cannibalizing women, noting Bollinger's claimed experience, and shared pictures with Bollinger of potential victims. Bollinger accepted Valle's offer, stating that, although he lived in England, it would be easy for him to travel to Valle and help him.
The two focused on Kimberly Sauer as their preferred target. Valle told Bollinger that she would be “the easiest” to abduct. App'x at 81. He explained that, because he knew her personally, he could “just show up at her home unannounced” without “alert[ing] her” in order to “knock her out ... and kidnap her.” Id. at 82. Valle suggested a “Labor Day cookout,” with Sauer “as the main course.” Id. at 86. Bollinger replied that he was “looking forward to it,” id., and would search for affordable plane tickets.
In multiple conversations over the course of a few weeks, Valle and Bollinger continued discussing the details of abducting and murdering Sauer. They considered logistical and practical concerns. They planned to use chloroform to incapacitate her, which Valle offered to make himself. They discussed stalking Sauer “in the evening,” id. at 82, noting that she lived alone, in a small house, far from family that might notice her absence. Valle offered to “do a dry run,” in which he would “show up randomly one day just being in the neighborhood.” Id. at 93. They also decided to cover the trunk of Valle's car in plastic bags and use “gloves and a new pair of really cheap, common trainers” that should be destroyed after the abduction, “probably by burning.” Id. at 95. They agreed that when buying materials they should use cash.
They also developed their plans to cook and eat Sauer. Valle and Bollinger debated the merits of roasting Sauer alive in an oven or over a rotisserie. They further discussed how she should be gagged and butchered. Valle promised to get a “brand new set of knives,” and Bollinger advised that they would need a “cleaver, not a saw.” Id. at 88.
As they plotted Sauer's abduction, Bollinger and Valle expressed excitement about their plan. Valle shared his happiness that Bollinger was “on board” and said that he loved imagining Sauer “asleep right now not having the slightest clue of what we have planned.” Id. at 90. When Bollinger replied that he was “looking forward to it so much,” Valle said that he could not wait to “see the look on her face when she wakes up naked and tied up.” Id. at 96. Valle noted that it was good that they were “brainstorm[ing],” because “everything[ ]needs to be perfect.” Id. at 93. They would “talk in great detail” about “every step,” suggested Valle, “beginning with the rag in her face.” Id.
Bollinger and Valle confirmed to each other their genuine intention to follow through on their plan. At one point, Bollinger asked Valle, “You WILL go through with this? I've been let down before. That's why i [sic] tend to work alone.” Id. at 91. Valle immediately replied “yes,” adding that Sauer would “never see it coming” and that he was anxious to kidnap and eat her. Id. Valle later stated that “kidnapping” Sauer and “getting away with it” was an “absolute truth.” Id. at 93.
Valle took what could be viewed reasonably as concrete steps to further his plan with Bollinger. He sent Bollinger a “blueprint” of their plot, id. at 100, entitled “Abducting and Cooking Kimberly,” id. at 267, which was consistent with what Valle and Bollinger had discussed up to that point. Valle searched the internet for Sauer's name, “how to kidnap someone,” id. at 383, “how to abduct a girl,” id. at 384, “[g]ood methods to kidnap someone,” id., “how to knock someone unconscious,” id. at 385, “how to chloroform a girl,” id., “how to tie someone up,” id. at 396, “most secure bondage,” id., and “how to hogtie a girl,” id. at 388. And Valle planned a trip, with his wife and young child, to Maryland, where Sauer lived. Valle contacted Sauer before the trip, and they planned a lunch while he and his family were in Maryland.
Before leaving for Maryland, Valle had further discussions with Bollinger. Valle explained that his lunch with Sauer would stimulate ideas for how to cook her, remarking that when he saw her, his mouth would “be watering.” Id. at 110. The two also discussed the lunch as a way to confirm their choice of Sauer as their preferred target. When Valle had a moment of indecision about whether to target different women instead of Sauer, Bollinger advised that they “[k]eep the others as spares” and “see how” the lunch with Sauer goes. Id. at 111. Bollinger further wished Valle “good luck,” id. at 116, cautioned him to act normally during the lunch—because he would be a “suspect” when Sauer “goes missing,” id. at 102—and asked him to report what he learned.
On the trip to Maryland, Valle text-messaged Sauer that he had driven past her place of work. Sauer found the text message strange and doubted that her other friends knew where she worked. Shortly after having lunch with Sauer, Valle wrote to Bollinger that Sauer “looked absolutely mouthwatering.” Id. at 117.
B. The Crime of Conspiracy
“To be complete, a conspiracy simply requires (i) an agreement about the object of the conspiracy, (ii) specific intent to achieve that object, and (iii) an overt act in furtherance of the agreement.” United States v. Wallace, 85 F.3d 1063, 1068 (2d Cir.1996). A defendant is guilty of conspiracy so long as he “agreed on the essential nature of the plan,” which need not accompany a fully-formed plot to constitute an illegal conspiracy. United States v. Eppolito, 543 F.3d 25, 47 (2d Cir.2008) (internal quotation marks omitted); see also United States v. Rosa, 17 F.3d 1531, 1543–44 (2d Cir.), cert. denied, 513 U.S. 879, 115 S.Ct. 211, 130 L.Ed.2d 140 (1994).
The evidence at trial, viewed in the light most favorable to the government, supported the jury's conclusion that Valle and another agreed to commit a kidnapping, intended at the time to do so, and performed at least one overt act in furtherance of their agreement, such as Valle's creation of the “blueprint” or meeting Sauer in Maryland. The jury could reach the conclusion that a conspiracy was formed even though, as it happened, Valle and Bollinger ultimately made no attempt to kidnap Sauer on Labor Day. A conspiratorial agreement is a distinct crime that is punishable regardless of whether the plan is later abandoned. See United States v. Jimenez Recio, 537 U.S. 270, 275, 123 S.Ct. 819, 154 L.Ed.2d 744 (2003) (explaining that a conspiracy “poses a threat to the public over and above the threat of the commission of the relevant substantive crime” (internal quotation marks omitted)).
C. The Jury's Reasonable Inferences
In spite of Valle and Bollinger's express agreement in their written communications to kidnap Sauer, the majority concludes that the jury acted irrationally when it found Valle guilty of conspiracy to kidnap. The majority reasons that, because Valle fantasized about cannibalism with others over the internet, it was irrational to believe that Valle and Bollinger meant what they said to each other. See Maj. Op. at 521–22. In my view, however, the majority fails to respect the reasonable inferences that the jury could have made in determining that Valle's plot with Bollinger—unlike his other online communications—was real.
In stark contrast to any of Valle's other online communications, Valle took actions in the real world that the jury could conclude were done to effectuate his plot with Bollinger. Valle visited Sauer in Maryland, drove past her workplace, and had lunch with her—a lunch that he and Bollinger discussed as a way to further their plans.1 But this was hardly, as the majority suggests, the “only meaningful difference,” id. at 522, between the plot concerning Sauer and the so-called “fantasy chats.” Only in Valle's conversation with Bollinger, for example, did the jury have evidence of a “blueprint” that Valle made of the plot.
The jury could also have found that the nature of Valle's conversations with Bollinger were distinct in tenor and tone. The detail of Valle's plans with Bollinger was unparalleled in his communications with others. And though Valle made allusions to being serious in certain of the supposed “fantasy chats,” at no point did he state so unequivocally, as he did to Bollinger, that it was an “absolute truth” that he and Bollinger would kidnap Sauer. App'x at 93.
Other evidence further supported the jury's conclusion that Valle's plot with Bollinger was more than mere fantasy. Valle's internet searches between the date of his first conversation with Bollinger and his lunch with Sauer were particularly relevant to his plan with Bollinger. The majority concedes that Valle's internet searches could have provided “relevant proof of intent,” but nevertheless dismisses them because, in its view, Valle's searches “occurred in a context of deep fantasy.” Maj. Op. at 519. This flawed analysis commits the fallacy of petitio principii (circular reasoning) because “what is to be proved is implicitly presumed as true in the premise.” Black's Law Dictionary 1329 (10th ed.2014); see also Adams v. Gould, Inc., 687 F.2d 27, 30 (3d Cir.1982) (explaining that “the fallacy of petitio principii” occurs when one “assum[es] the conclusion”), cert. denied, 460 U.S. 1085, 103 S.Ct. 1777, 76 L.Ed.2d 348 (1983); Nico v. Comm'r of Internal Revenue, 565 F.2d 1234, 1238 (2d Cir.1977) (“this court by all means eschews petitio principii”). The majority concludes that Valle was engaged in fantasy based on its reading of his written communications and then finds his internet searches consistent with this preordained conclusion. The jury, however, may have considered Valle's internet searches in deciding that very issue—whether Valle was engaged only in fantasy.
The jury could have also considered Valle's post-arrest statement to a government agent that Bollinger was “more serious” about their discussions than other individuals with whom Valle communicated. Trial Tr. at 1031.
The majority opinion takes pains to conjure innocent explanations for this evidence. Perhaps Valle's lunch with Sauer was just a normal social interaction. See Maj. Op. at 521–22. Maybe Valle withheld from Bollinger Sauer's address because he did not intend to actually harm her. See id. Possibly Valle conducted internet searches on how to kidnap someone simply because he had a general interest in crime rather than criminal intent.2 See id. at 519–20. These are perspectives on the evidence that the jury conceivably could have found persuasive but did not. It was the jury's sole prerogative to decide how to interpret the evidence presented, and the majority's notions of how the evidence should instead have been interpreted are incompatible with our obligation to respect the jury's fact-finding role by viewing the evidence in the light most favorable to the government and drawing all reasonable inferences in its favor. See United States v. Aguiar, 737 F.3d 251, 265 (2d Cir.2013) (“We agree that much of the evidence could be read to have an innocent meaning, but when the evidence raises two permissible inferences then we must resolve such conflicts in favor of the prosecution.”), cert. denied, ––– U.S. ––––, 135 S.Ct. 400, 190 L.Ed.2d 290 (2014); United States v. Friedman, 998 F.2d 53, 56 (2d Cir.1993) (explaining that “[i]t is not for [the court] to weigh ... competing inferences and explanations” to ascertain “which explanation [of defendant's conduct] is more likely”); see also, e.g., United States v. Pavulak, 700 F.3d 651, 670 (3d Cir.2012) (rejecting defendant's contention that his requests to see a child naked were “facetious ‘banter’ ” and “fantasies,” because it was not for the court “to weigh the evidence”), cert. denied, ––– U.S. ––––, 133 S.Ct. 2047, 185 L.Ed.2d 904 (2013); United States v. Dwinells, 508 F.3d 63, 74 (1st Cir.2007) (acknowledging that defendant's response to charges of attempted enticement of a minor—“that he was merely role-playing and thought that the communications were mutually entertained fantasies”—was “plausible” and “buttressed by [his] persistent dodging of suggestions that he and his correspondents meet,” but concluding that “the government's theory of the case ... also was plausible,” and that “[w]hen the record is fairly susceptible to two competing scenarios, the choice between those scenarios ordinarily is for the jury”), cert. denied, 554 U.S. 922, 128 S.Ct. 2961, 171 L.Ed.2d 892 (2008).
For each assessment of the evidence made by the majority in contravention of the jury's finding of guilt, there is a contrary light through which the evidence can be viewed—in fact, must be viewed—that supports the jury's conclusion.
The majority notes, for instance, that in his conversations with Bollinger, Valle lied about certain seemingly important facts (such as where he lived, whether he knew Sauer's address, and whether he owned a secluded mountain house). See Maj. Op. at 520–22. The majority similarly points out that Valle's “blueprint” of Sauer's abduction listed inaccurately much of Sauer's identifying information. See id. at 521. But Valle's misstatements hardly demonstrate a lack of intent. For instance, the jury could have rationally inferred from the evidence that, while the plan was still developing, Valle misled Bollinger because he did not fully trust him. Valle's “blueprint” obscures information about Sauer that could be used to find her easily, such as her last name, but other information in the document is, by contrast, accurate (e.g., Sauer's picture, marital status, lack of tattoos). Bollinger, in fact, expressly assumed that mistrust was Valle's reason for not sharing certain information; when Valle claimed not to know Sauer's address, Bollinger remarked, “not like I'll get there a day early!” App'x at 101. And Valle's lies about what he possessed to facilitate the crime—such as a secluded mountain house—were in response to Bollinger's concerns about the plan's effectiveness; the jury could have reasoned that Valle fibbed to ensure Bollinger's interest. Valle would not be the first defendant to mislead a coconspirator, and his misstatements do not negate the jury's verdict. E.g., United States v. Gersh, 328 F.2d 460, 462 (2d Cir.) (stating that there was “no less a meeting of the minds” because coconspirators decided to deceive each other), cert. denied, 377 U.S. 992, 84 S.Ct. 1919, 12 L.Ed.2d 1045 (1964); cf. United States v. Farley, 607 F.3d 1294, 1335 (11th Cir.) (rejecting defendant's assertion that his lack of intent was demonstrated by lies to his coconspirator about details in their plot to sexually abuse a minor), cert. denied, 562 U.S. 945, 131 S.Ct. 369, 178 L.Ed.2d 238 (2010).
The majority discounts Valle's internet searches as demonstrating merely an “interest” in criminal activity. See Maj. Op. at 519. But the jury could certainly reasonably deduce that Valle's inquiries into how to incapacitate, bind, and kidnap women—all conducted contemporaneously with his conversations with Bollinger about abducting Sauer—provided a strong indication of criminal intent.
The majority likewise supposes that the improbability of Valle and Bollinger's plot suggests that it was no more than fantasy. Valle and Bollinger met on a sexual fetish website, knew almost nothing about each other, barely discussed the logistics of Bollinger traveling thousands of miles to assist Valle, developed a “blueprint” that was “no more detailed than ... Valle's Internet chats,” id. at 521–22, and let their target date for Sauer's abduction pass without comment. Yet Valle and Bollinger explicitly stated that their plans were incomplete and needed more work. When Valle sent Bollinger the “blueprint,” in fact, he acknowledged that “obviously a lot has to be added” because they were “in the beginning stages.” App'x at 100. And throughout their discussions of Sauer, Valle and Bollinger contemplated additional or alternative targets. The evidence supports the inference that Valle and Bollinger agreed on a basic framework and set an aspirational date but then focused on other targets or plans that they later found more appealing.
The majority also suggests that the jury lacked evidence of Bollinger's criminal intent. See Maj. Op. at 522–23. To the contrary, the only evidence at trial as to Bollinger was his written conversations with Valle, in which he unequivocally established intent by expressly agreeing to help Valle kidnap Sauer. The majority's analysis as to why the jury should have disbelieved Valle's intent is irrelevant to the question of Bollinger's intent. Whereas the majority points to Valle's other online conversations to contend that his statements to Bollinger were feigned, no such evidence existed as to Bollinger.
D. The Jury's Verdict Was Sufficiently Supported by the Evidence
As judges reviewing the sufficiency of the evidence in a criminal trial, given the grave importance of the event, it can be tempting to view the evidence in the light most convincing to us, to discount plausible factual inferences that we find unpersuasive, and to conclude that if we do not believe that the evidence proved guilt beyond a reasonable doubt, then no rational juror could have so found. That, however, is simply not the standard to which we are bound, see Jackson, 443 U.S. at 318–19, 99 S.Ct. 2781; United States v. Josephberg, 562 F.3d 478, 488 (2d Cir.) (“The weight of the evidence is a matter for argument to the jury, not a ground for reversal on appeal.” (internal quotation marks omitted)), cert. denied, 558 U.S. 965, 130 S.Ct. 397, 175 L.Ed.2d 302 (2009), and we must be diligent not to overreach into the jury's domain, cf. United States v. Butler, 297 U.S. 1, 79, 56 S.Ct. 312, 80 L.Ed. 477 (1936) (Stone, J., dissenting) (“[T]he only check upon our own exercise of power is our own sense of self-restraint.”). Our “deference to the jury's findings is especially important” in a conspiracy case, “because a conspiracy by its very nature is a secretive operation, and it is a rare case where all aspects of a conspiracy can be laid bare in court with the precision of a surgeon's scalpel.” United States v. Santos, 541 F.3d 63, 70 (2d Cir.2008) (internal quotation marks omitted), cert. denied, 555 U.S. 1122, 129 S.Ct. 960, 173 L.Ed.2d 150 (2009).
Here, I fear the majority treads beyond the strict limitations of our review by improperly weighing competing inferences and explanations of the evidence—as if deciding a motion for a new trial instead of reviewing a judgment of acquittal—rather than viewing it, as we must, “in its totality,” Hawkins, 547 F.3d at 70 (internal quotation marks), and in the government's favor, see Tibbs v. Florida, 457 U.S. 31, 45, 102 S.Ct. 2211, 72 L.Ed.2d 652 (1982) (recognizing “lower limit on an appellate court's definition of evidentiary sufficiency” compared to evidentiary weight); see also, e.g., United States v. Autuori, 212 F.3d 105, 120 (2d Cir.2000) (affirming grant of motion for new trial while reversing judgment of acquittal).
All that was required to find Valle guilty of conspiracy was proof of one agreement with one coconspirator to kidnap one target, as well as one overt act. See Wallace, 85 F.3d at 1068. Even if the government's evidence was insufficient in respect of each of the other four alleged kidnapping conspiracies, see Maj. Op. at 512–13, Valle's conversations with Bollinger about kidnapping and cannibalizing Sauer, together with Valle's overt acts, such as his trip to Maryland, were enough for the jury to rationally find that Valle committed the crime of conspiracy to kidnap, see United States v. Berger, 224 F.3d 107, 113 (2d Cir.2000) (“[T]he government need[ ] only [ ] prove agreement on one of the objectives charged in the indictment in order to establish that a conspiracy existed.”).
Of course, a jury can never determine with complete certainty what a criminal defendant was thinking at a particular moment. But here the jury had sufficient evidence to conclude, beyond a reasonable doubt, that Valle and Bollinger meant exactly what they said to each other. See United States v. Rowe, 56 F.2d 747, 749 (2d Cir.) (L. Hand, J.) (“A jury alone could say whether in such a setting the hearer ought to have understood that the utterances were not to be taken at their face....”), cert. denied, 286 U.S. 554, 52 S.Ct. 579, 76 L.Ed. 1289 (1932). Valle and Bollinger targeted Sauer because she would be easy to kidnap, spent hours plotting her abduction, and repeatedly stated that they were serious—and excited—about their plan. Their words were corroborated by Valle's trip to Maryland, contemporaneous internet searches, blueprint for abducting and cooking Sauer, and post-arrest assertion that Bollinger was “more serious” than others with whom Valle communicated.3 The jury could reasonably deduce that the tenor of Valle's conversations with Bollinger, together with his real-world actions, proved beyond a reasonable doubt that the plot to kidnap Sauer was no fantasy. We cannot second-guess the jury's rational conclusion. See Jackson, 443 U.S. at 319, 99 S.Ct. 2781; United States v. Allied Stevedoring Corp., 241 F.2d 925, 930 (2d Cir.) (L.Hand, J.) (“What weight the jury should give to [the] evidence was for them, and them alone, provided that it satisfied their minds beyond any fair doubt.”), cert. denied, 353 U.S. 984, 77 S.Ct. 1282, 1 L.Ed.2d 1143 (1957).
I would thus vacate the District Court's judgment of acquittal in respect of Count I and remand for the District Court to consider Valle's alternative arguments for acquittal, which were raised below but did not form the basis for the District Court's decision.
II. Computer Fraud and Abuse Act (Count II)
Valle was convicted of conducting a computer search that exceeded his authorized access to a federal law enforcement database, in violation of the Computer Fraud and Abuse Act, 18 U.S.C. § 1030. Because I agree with the District Court that Valle's conduct is proscribed by the plain language of the CFAA, I would affirm his conviction.
The CFAA imposes criminal penalties on anyone who “intentionally accesses a computer without authorization or exceeds authorized access, and thereby obtains ... information from any department or agency of the United States.” 18 U.S.C. § 1030(a)(2)(B). The phrase “exceeds authorized access” is defined as “access[ing] a computer with authorization” to obtain “information in the computer that the accesser is not entitled so to obtain.” Id. § 1030(e)(6). The CFAA therefore protects information belonging to the United States both from those who lack any authorization to access the computer system and from those, like Valle, who did not comply with restrictions on their authorized access.
The evidence at trial established that, by entering the name “Maureen Hartigan” into a federal law enforcement database, Valle obtained information that he was “not entitled so to obtain.” See United States v. Valle, 301 F.R.D. 53, 109–10 (S.D.N.Y.2014). Valle was instructed repeatedly that he had permission to use his credentials to access the federal National Crime Information Center (“NCIC”) database only when “required to do so in the course of [his] official duties and responsibilities” as a police officer, and that there were “no exceptions to this policy.” App'x at 61–62. Valle was warned that accessing law enforcement databases for non-official purposes was improper and that the penalties for so doing included termination and prosecution. And Valle concedes that he had no legitimate law enforcement purpose when he queried the database for the name “Maureen Hartigan.” See Valle, 301 F.R.D. at 110 (“It is undisputed that Valle had no law enforcement purpose for querying Hartigan's name....”).
The majority nonetheless holds that, because Valle possessed the technical credentials to access the NCIC database and query Hartigan's name, he did not exceed his authorized access by so doing. See Maj. Op. at 523–24, 527–28. In reaching this result, the majority discovers ambiguity in the statutory language where there is none. Under the plain language of the statute, Valle exceeded his authorized access to a federal database in violation of the CFAA.
Statutory construction must “begin with the language employed by Congress and the assumption that the ordinary meaning of that language accurately expresses the legislative purpose.” Shi Liang Lin v. U.S. Dep't of Justice, 494 F.3d 296, 305 (2d Cir.2007) (en banc) (internal quotation marks omitted), cert. denied, 553 U.S. 1053, 128 S.Ct. 2472, 171 L.Ed.2d 766 (2008). “Congress says in a statute what it means and means in a statute what it says there.” Id.
As the majority concedes, a preponderance of our sister circuits interpreting the term “exceeds authorized access” have concluded that the statute unambiguously encompasses conduct of the type engaged in by Valle. See Maj. Op. at 525; see also United States v. Rodriguez, 628 F.3d 1258, 1263 (11th Cir.2010), cert. denied, 563 U.S. 966, 131 S.Ct. 2166, 179 L.Ed.2d 946 (2011); United States v. John, 597 F.3d 263, 270–73 (5th Cir.2010); Int'l Airport Ctrs., LLC v. Citrin, 440 F.3d 418, 420–21 (7th Cir.2006); EF Cultural Travel BV v. Explorica, Inc., 274 F.3d 577, 581–84 (1st Cir.2001). The Eleventh Circuit, in United States v. Rodriguez, affirmed the conviction of a Social Security Administration employee who obtained personal information about former girlfriends and other women from federal databases that he was authorized to use only for business reasons. 628 F.3d at 1260–63. Rodriguez argued that he did not violate Section 1030(a)(2)(B) because he accessed databases that he was authorized to use as a Social Security Administration employee. See id. at 1263. The Eleventh Circuit rejected his argument, explaining that Rodriguez clearly “exceed[ed] his authorized access” because “his access of the victims' personal information was not in furtherance of his duties” as an employee. Id.
Similarly, in United States v. John, the Fifth Circuit found that a bank employee exceeded her authorized access when, in order to facilitate fraud, she accessed confidential customer information in contravention of her employer's restrictions on computer use. 597 F.3d at 270–73. There, as here, the defendant was aware that her employer's “official policy,” which was reiterated in training programs she attended, “prohibited misuse of the company's internal computer systems and confidential ... information.” Id. at 272.
The majority concludes that these courts “looked only at the culpable behavior of the defendants before them, and failed to consider the effect on millions of ordinary citizens caused by the statute's unitary definition of ‘exceeds authorized access.’ ” Maj. Op. at 527 (internal quotation marks omitted). It notes concerns, articulated by two of our sister circuits, about the potentially expansive scope of the CFAA.4 See id. at 524–28 (discussing United States v. Nosal, 676 F.3d 854, 859–63 (9th Cir.2012) (en banc) (observing that a broad construction of the CFAA “would make criminals of large groups of people who have little reason to suspect they are committing a federal crime” (internal quotation marks omitted)), and WEC Carolina Energy Sols., LLC v. Miller, 687 F.3d 199, 206 (4th Cir.2012), cert. dismissed, ––– U.S. ––––, 133 S.Ct. 831, 184 L.Ed.2d 645 (2013)).
The majority opinion, apparently without irony, concludes that giving effect to the plain language of the statute would somehow “place us in the position of [the] legislature.” Id. at 527. But where, as here, the statute's language is plain and unambiguous, the “sole function of the courts is to enforce it according to its terms.” United States v. Ron Pair Enters., Inc., 489 U.S. 235, 241, 109 S.Ct. 1026, 103 L.Ed.2d 290 (1989); accord United States v. DiCristina, 726 F.3d 92, 96 (2d Cir.2013), cert. denied, ––– U.S. ––––, 134 S.Ct. 1281, 188 L.Ed.2d 299 (2014). It may well be that the CFAA sweeps broadly. But such is a matter for policy debate, see United States v. Rodgers, 466 U.S. 475, 483, 104 S.Ct. 1942, 80 L.Ed.2d 492 (1984) (“Resolution of the pros and cons of whether a statute should sweep broadly or narrowly is for Congress.”), and the Congress is free to amend the statute if it chooses, see Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 576, 102 S.Ct. 3245, 73 L.Ed.2d 973 (1982) (“The remedy for any dissatisfaction with the results in particular cases lies with Congress and not with this Court. Congress may amend the statute; we may not.”). “Any other conclusion, while purporting to be an exercise in judicial restraint, would trench upon the legislative powers vested in Congress by Art. I, § 1, of the Constitution.” Dep't of Housing & Urban Dev. v. Rucker, 535 U.S. 125, 134–35, 122 S.Ct. 1230, 152 L.Ed.2d 258 (2002) (internal quotation marks omitted); cf. Dennis v. United States, 341 U.S. 494, 526, 71 S.Ct. 857, 95 L.Ed. 1137 (1951) (Frankfurter, J., concurring) (“[W]e must scrupulously observe the narrow limits of judicial authority even though self-restraint is alone set over us. Above all we must remember that this Court's power of judicial review is not an exercise of the powers of a super-Legislature.”).
See also Principal Deputy Assistant Attorney General David Bitkower Delivers Keynote Address at George Washington Law Review Symposium entitled “Hacking into the Computer Fraud and Abuse Act,” U.S. Dep't of Justice (Nov. 6, 2015), http://www.justice.gov/opa/speech/principal-deputy-assistant-attorneygeneral-david-bitkower-delivers-keynote-address (describing proposed amendment to CFAA to clarify definition of “exceeds authorized access”); Cyber Crime: Modernizing Modernizing our Legal Framework for the Information Age: Hearing Before the Subcommittee on Crime and Terrorism of the Senate Committee on the Judiciary, 114th Cong. (2015) (statement of David M. Bitkower, Deputy Assistant Attorney General, Department of Justice), available at http://www.judiciary.senate.gov/imo/media/doc/07–08–15% 20Bitkower% 20Testimony.pdf (describing proposed amendments to limit breadth of CFAA).
The majority invokes the rule of lenity in support of its view. See Maj. Op. at 526–28. The doctrine is a “rule of last resort.” Oppedisano v. Holder, 769 F.3d 147, 153 (2d Cir.2014) (internal quotation marks omitted), cert. denied, ––– U.S. ––––, 136 S.Ct. 211, 193 L.Ed.2d 129 (2015); see also Reno v. Koray, 515 U.S. 50, 65, 115 S.Ct. 2021, 132 L.Ed.2d 46 (1995) (“The rule of lenity applies only if, after seizing everything from which aid can be derived, we can make no more than a guess as to what Congress intended.” (internal citations and quotation marks omitted)). It “serves to aid the court in interpreting a criminal statute only if there is an ambiguity,” and it cannot be “used to narrow a statute that has an unambiguously broad thrust.” United States v. Litchfield, 986 F.2d 21, 22 (2d Cir.1993) (per curiam); see also Abramski v. United States, ––– U.S. ––––, 134 S.Ct. 2259, 2272 n. 10, 189 L.Ed.2d 262 (2014) (“The dissent would apply the rule of lenity here because the statute's text, taken alone, permits a narrower construction, but we have repeatedly emphasized that is not the appropriate test.”).
The majority concludes that the “sharp division” among our sister circuits means that whether the statute proscribes Valle's conduct cannot be resolved simply by looking at the text of the statute, requiring us to “turn to the legislative history and motivating policies for further guidance.” Maj. Op. at 524–25. To the contrary, however, the Supreme Court has explained that “[a] statute is not ambiguous for purposes of lenity merely because there is a division of judicial authority over its proper construction.” Reno, 515 U.S. at 64–65, 115 S.Ct. 2021 (internal quotation marks omitted); see also DiCristina, 726 F.3d at 104 (“A statute is not ‘ “ambiguous” for purposes of lenity merely because it [i]s possible to articulate a construction more narrow than that urged by the Government.’ ” (quoting Moskal v. United States, 498 U.S. 103, 108, 111 S.Ct. 461, 112 L.Ed.2d 449 (1990))).
Therefore, “[w]here statutory ... provisions unambiguously cover the defendant's conduct,” as Section 1030(a)(2)(B) clearly proscribes Valle's conduct here, the rule of lenity “does not come into play.” Litchfield, 986 F.2d at 22; see, e.g., DePierre v. United States, 564 U.S. 70, 131 S.Ct. 2225, 2237, 180 L.Ed.2d 114 (2011) (explaining that the rule of lenity was inapplicable “[b]ecause the statutory text allows us to make far more than a guess as to what Congress intended” (internal quotation marks omitted)); Rodgers, 466 U.S. at 484, 104 S.Ct. 1942 (concluding that criminal statute was “not sufficiently ambiguous ... to permit the rule [of lenity] to be controlling”); see also Yates v. United States, ––– U.S. ––––, 135 S.Ct. 1074, 1098–99, 191 L.Ed.2d 64 (2015) (Kagan, J., dissenting) (“Lenity offers no proper refuge from [a] straightforward (even though capacious) construction.”). For the same reason, I also have no occasion to refer to the statute's legislative history. See United States v. Woods, ––– U.S. ––––, 134 S.Ct. 557, 567 n. 5, 187 L.Ed.2d 472 (2013) (Scalia, J.) (“Whether or not legislative history is ever relevant, it need not be consulted when, as here, the statutory text is unambiguous.”); Puello v. Bureau of Citizenship & Immigration Servs., 511 F.3d 324, 331 (2d Cir.2007) (Katzmann, J.) (“Because we believe the language in the statute is unambiguous, we need not examine legislative history to divine the statute's meaning.”).
I emphasize that I take no position on the applicability of Section 1030(a)(2)(B) in other circumstances or the scope or validity of other provisions of the CFAA, which are not at issue here. Because I find that Section 1030(a)(2)(B) of the CFAA clearly proscribes Valle's conduct, I would affirm.
The majority apparently rules not on the issue presented on this appeal—whether Valle's conduct is proscribed by the CFAA—but on the application of this statute to other circumstances not implicated here. See Maj. Op. at 528 (explaining that, “[w]hatever the apparent merits of imposing criminal liability may seem to be in this case,” the majority is “constru[ing] the statute” for the purposes of “many other situations”). Because the majority's ruling might “impact[ ] many more people than Valle,” it rejects the government's reasonable proposal that we let concerns about “the risk of criminalizing ordinary behavior” be addressed in due course when raised “by individuals actually affected by the provision at issue.” Id. (internal quotation marks omitted). In my view, the majority's approach is unwise and improper. See Costello v. INS, 311 F.2d 343, 348 (2d Cir.1962) (“[W]e think the exercise of a proper judicial restraint makes it undesirable for us to reach out and now decide more than the circumstances of this case require us to decide.”), rev'd on other grounds, 376 U.S. 120, 84 S.Ct. 580, 11 L.Ed.2d 559 (1964).
I would vacate the District Court's judgment of acquittal in respect of Count I, affirm the District Court's judgment of conviction in respect of Count II, and remand for further proceedings. Therefore, I respectfully dissent.