United States v. Adrian Grisanti

In the United States Court of Appeals For the Seventh Circuit ____________________ Nos. 18-2993 & 19-1576 UNITED STATES OF AMERICA, Plaintiff-Appellee, v. ADRIAN GRISANTI, Defendant-Appellant. ____________________ Appeals from the United States District Court for the Southern District of Indiana, New Albany Division. No. 4:16-cr-00018-TWP-VTW-1 — Tanya Walton Pratt, Judge. ____________________ ARGUED OCTOBER 2, 2019 — DECIDED NOVEMBER 22, 2019 ____________________ Before BAUER, RIPPLE, and HAMILTON, Circuit Judges. HAMILTON, Circuit Judge. Appellant Adrian Grisanti was convicted of child-pornography offenses and destruction of evidence. On appeal, he challenges the denial of his motion to suppress evidence and the length of his sentence. We affirm on both issues. We have already held that the good-faith exception applies to the same warrant at issue in this case, which authorized the use of a sophisticated technique to identify users of a child-pornography website. See United States 2 Nos. 18-2993 & 19-1576 v. Kienast, 907 F.3d 522, 529 (7th Cir. 2018). Grisanti’s reasons for reconsidering Kienast are not persuasive. Also, his sentence was not unreasonable and the district court did not make any procedural error. I. Factual Background and Procedural History The Federal Bureau of Investigation gained control of a child-pornography website called “Playpen.” The FBI kept Playpen running for two weeks from a server in Newington, Virginia, to locate people who distributed and viewed child pornography on the site. Because Playpen allowed visitors to use it anonymously, the FBI applied to a magistrate judge in the Eastern District of Virginia for a warrant authorizing the use of a “Network Investigative Technique,” or “NIT,” to identify the site’s users. When a user logged into Playpen, the NIT installed malware on the user’s computer and relayed identifying information about that computer back to the FBI’s server in Virginia. An affidavit supporting the warrant application explained this to the magistrate judge. Exactly where these searches would occur was not quite as clear. The application said that the property to be searched was “located in the Eastern District of Virginia.” The application’s “Place to be Searched” addendum stated, though, that the NIT would be “deployed” on a server “located at a government facility in the Eastern District of Virginia” to obtain information from “activating computer[s]”—those of “any user” who logged into Playpen. And the supporting affidavit added that the NIT “would cause an activating computer— wherever located—to send” information to the Virginia server. Based on these representations, the magistrate judge issued a warrant authorizing the use of the NIT to search Nos. 18-2993 & 19-1576 3 property “located in the Eastern District of Virginia” to obtain information from the computers of Playpen users. When Grisanti logged into Playpen from his work computer in Indiana, the NIT malware was installed and sent identifying information to the FBI. Using that information, the FBI obtained additional search warrants in Indiana and found evidence of child pornography on Grisanti’s work computer. Before the FBI could complete its investigation, however, Grisanti learned of the inquiry. He destroyed the hard drive and a flash drive. He was charged in the Southern District of Indiana with destruction of evidence and several child-pornography offenses. See 18 U.S.C. §§ 1519 , 2252A(a)(2)(A), & 2252A(a)(5)(B). Grisanti moved to suppress all evidence obtained as a result of the NIT warrant. Judge Pratt agreed with him that the warrant was invalid because the magistrate judge had exceeded her jurisdiction by authorizing searches outside of the Eastern District of Virginia, in violation of the Federal Magistrates Act, 28 U.S.C § 636(a)(1), and the version of Federal Rule of Criminal Procedure 41(b) in effect when the warrant was issued in 2015. Judge Pratt denied the motion to suppress, however, concluding that the FBI agents had relied on the warrant in good faith. 1 A jury found Grisanti guilty of all charges. The Presentence Report proposed a Sentencing Guideline range of 108 to 135 months in prison. Judge Pratt later adopted that calculation without objection. In Grisanti’s sentencing memoran1 As noted in Kienast, 907 F.3d at 527 n.1, Rule 41 was amended in 2016 to permit magistrate judges to issue warrants like the NIT warrant here. See Fed. R. Civ. P. 41(b)(6)(A). 4 Nos. 18-2993 & 19-1576 dum, he requested a 78-month sentence, arguing that he had an “addiction” to child pornography and needed treatment. He also cited the “Child Pornography Offender Risk Tool,” the “Correlates of Admission of Sexual Interest in Children” assessment, and other research to argue that he is unlikely to recidivate because he is “white, employed, and had no previous criminal history,” and had not committed any so-called “contact offenses.” 2 At the sentencing hearing, Grisanti’s attorney argued that Grisanti had “compartmentalized” his criminal behavior and was otherwise a “good, loving husband” and “caring employee.” The judge raised her concern that Grisanti had defended himself by blaming others and still failed to take full responsibility. Even though Grisanti accepted that he had “an addiction or a criminal problem,” the judge noted, he would not be able to “get better” if he refused to accept that he had an “issue” that “needs treatment”—he would be “just like” another defendant whom the judge had sentenced earlier that day for his second conviction for child pornography. The judge also questioned Grisanti’s argument about low risk of recidivism and specifically whether race had “anything to do with being a child pornographer.” The judge asked further if a psychologist had assessed Grisanti. Defense counsel answered that Grisanti had not been evaluated and agreed with 2 See Angel Wyatt Eke et al., Scoring Guide for the Child Pornography Offender Risk Tool (CPORT): Version 2, ResearchGate (2018); Angel Wyatt Eke et al., Scoring Guide for the Child Pornography Offender Risk Tool (CPORT): Version 2, ResearchGate (2018); Michael Seto et al., Contact Sexual Offending by Men with Online Sexual Offenses, 23 Sexual Abuse: A Journal of Research & Treatment 124 (2011). Nos. 18-2993 & 19-1576 5 the judge that, “for him to be in a position to not reoffend, yes, he has to be prepared to get help and treatment.” The court then heard the remainder of counsel’s arguments, Grisanti’s allocution, and the government’s argument for a sentence of 135 months, at the high end of the guideline range. The judge sentenced Grisanti to 120 months in prison. In explaining the decision, she emphasized the seriousness of the crime: Grisanti possessed more than 600 images of child pornography—some involving prepubescent children—and then destroyed the evidence to thwart the investigation. He knew his actions were wrong, the judge continued, but if he had a problem, he never sought treatment for it and instead blamed others when he was caught. Finally, the judge agreed with Grisanti that he “could fulfill much of his untapped potential and move on with his life” if he got some treatment. II. The Motion to Suppress and the Good-Faith Exception The NIT warrant at issue here has led to many prosecutions and has been attacked by defendants across the country. Grisanti joins them, challenging the denial of his motion to suppress the evidence obtained through its use. He acknowledges that in United States v. Kienast, 907 F.3d 522, 529 (7th Cir. 2018), we held that the good-faith exception applies to agents who relied on this very warrant. Ten other circuits have agreed with that conclusion: United States v. Levin, 874 F.3d 316 , 321–24 (1st Cir. 2017); United States v. Eldred, 933 F.3d 110 , 118–21 (2d Cir. 2019); United States v. Werdene, 883 F.3d 204 , 215–18 (3d Cir. 2018); United States v. McLamb, 880 F.3d 685 , 689–91 (4th Cir. 2018); United States v. Ganzer, 922 F.3d 579 , 587–90 (5th Cir. 2019); United States v. Moorehead, 912 F.3d 963 , 967–71 (6th Cir. 2019); United States v. Horton, 863 F.3d 1041 , 1051–52 (8th Cir. 2017); United States v. Henderson, 906 F.3d — 6 — Nos. 18-2993 & 19-1576 1109, 1117–20 (9th Cir. 2018); United States v. Workman, 863 F.3d 1313 , 1317–21 (10th Cir. 2017); United States v. Taylor, 935 F.3d 1279, 1282 (11th Cir. 2019). Grisanti argues, however, that the good-faith exception should be deemed categorically inapplicable to warrants that are issued “without jurisdiction” and thus, he contends, “void ab initio.” In Kienast, 907 F.3d at 528 , we found that this argument is incompatible with the Supreme Court’s decision in Herring v. United States, 555 U.S. 135 (2009), which applied the good-faith exception to a search based on a warrant that already had been recalled. Accord, Eldred, 933 F.3d at 120 (applying Herring to the NIT warrant); Werdene, 883 F.3d at 216 (same); Ganzer, 922 F.3d at 587 (same); Moorehead, 912 F.3d at 969 (same); Horton, 863 F.3d at 1051 (same); Henderson, 906 F.3d at 1119 (same); Workman, 863 F.3d at 1318 & n.1 (same); Taylor, 935 F.3d at 1290–91 (same). Even if the magistrate judge lacked jurisdiction or some other degree of authority to issue a warrant that reached beyond the Eastern District of Virginia, we must consider whether the good-faith exception applies. “[W]hen an officer acting with objective good faith has obtained a search warrant from a judge or magistrate and acted within its scope,” the good-faith exception generally applies even if the warrant turns out to be invalid. United States v. Leon, 468 U.S. 897, 920 (1984). Suppression remains an appropriate remedy if the officer misled the issuing judge with knowingly false information or reckless disregard of the truth, or if a reasonable officer would know from the face of the warrant that it was invalid or that the search goes beyond its scope. Id. at 923 ; see also Groh v. Ramirez, 540 U.S. 551 , 564– Nos. 18-2993 & 19-1576 7 65 & n.8 (2004) (denying officer qualified immunity where warrant omitted items to be seized). 3 In Kienast, the defendants argued that the good-faith exception should not apply because a well-trained officer would have known that the magistrate judge lacked authority to authorize searches of computers across the country and therefore would have recognized that the NIT warrant was facially invalid. We rejected that argument. 907 F.3d at 528 . Grisanti concedes that if the warrant invalidly authorized a nationwide search, then the magistrate judge is to blame. He argues, though, that the FBI is “not faultless” for having sought and executed such an expansive warrant “from a magistrate judge with limited territorial jurisdiction.” This argument, however, is no different from that presented in and rejected by Kienast. Characterizing the extent of the magistrate judge’s power to issue the NIT warrant as an unsettled and difficult question, we concluded in Kienast that the FBI could have reasonably believed that the magistrate judge had the requisite authority. Specifically, because Rule 41(b)(4) permits a magistrate judge to authorize the installation of a “tracking device” within her district to track movement outside the district, so too might a magistrate judge be able to permit an electronic search of property outside the district. Id. at 529 . Suppressing the evidence from the NIT warrant would be inappropriate because penalizing the officer for the magistrate’s arguable error rather than his own “cannot logically contribute to the 3Leon also held that the good-faith exception is unavailable when the magistrate judge “wholly abandoned” her neutral judicial role or when the warrant is based on an affidavit “so lacking in indicia of probable cause as to render official belief in its existence entirely unreasonable.” 468 U.S. at 923 (citations omitted). Grisanti does not rely on these grounds. 8 Nos. 18-2993 & 19-1576 deterrence of Fourth Amendment violations.” Kienast, 907 F.3d at 528 , quoting Leon, 468 U.S. at 921 .4 Grisanti raises other grounds for declining to apply the good-faith exception, which we did not address in Kienast. First, he argues, the FBI obtained the warrant in bad faith because the affidavit assured the magistrate judge that the “property” to be searched was “located in the Eastern District of Virginia,” though the FBI planned to search computers anywhere in the world. Without dwelling on the rather slippery question of just where the searches would take place, we find that the affidavit sufficiently informed the magistrate judge that the FBI would be obtaining identifying information from computers outside her district—“wherever located.” See McLamb, 880 F.3d at 690–91; Horton, 863 F.3d at 1051–52; Taylor, 935 F.3d at 1292 ; see also United States v. Spears, 673 F.3d 598 , 605–07 (7th Cir. 2012) (applying good-faith exception where omissions and inconsistencies were immaterial). The difference between the affidavit and a more detailed description of how the software would work does not allow an inference that the agents intentionally misrepresented or recklessly omitted material information to mislead the magistrate judge. See United States v. Daniels, 906 F.3d 673, 677 (7th Cir. 2018); Leon, 468 U.S. at 923 ; see also Taylor, 935 F.3d at 1292 (holding that application and 4 Kienast did not actually decide whether Rule 41(b)(4) confers this power on magistrate judges, but all circuits that have answered the question have concluded that issuance of the warrant by the magistrate judge, as opposed to a district judge, violated both Rule 41 (before the 2016 amendment) and the Fourth Amendment. See Werdene, 883 F.3d at 211– 14; Horton, 863 F.3d at 1046–49; Henderson, 906 F.3d at 1113–17; Taylor, 935 F.3d at 1286–88. As in Kienast, however, we need not decide that underlying issue. Nos. 18-2993 & 19-1576 9 affidavit for NIT warrant “sufficiently disclosed bounds of intended search”). If the magistrate judge had had concerns about geography, the application made clear that the FBI sought permission to use the equipment in Virginia to obtain information from users’ computers wherever they were located. The application also provided ample information to have prompted further questions before the NIT warrant was issued if geographic limits had been a concern. Second, Grisanti argues that the government cannot rely on the good-faith exception because the agents knowingly exceeded the scope of the warrant, which purportedly authorized searches only within the Eastern District of Virginia. The good-faith exception does not apply to a search that clearly exceeds the scope of a warrant. Leon, 468 U.S. at 921–22; see also Messerschmidt v. Millender, 565 U.S. 535, 555 (2012) (examining, in qualified-immunity context, officer’s good faith in construing validity and scope of warrant); Groh, 540 U.S. at 561 n.4. Although the NIT warrant did not specifi- cally authorize searches outside the Eastern District of Virginia, it permitted agents to obtain information from any computer used to log into Playpen. On this basis, the First, Second, Third, and Ninth Circuits have all held that a reasonable agent could rely on the NIT warrant as permitting the use of the software and equipment in Virginia to obtain information from computers outside the magistrate judge’s district. Levin, 874 F.3d at 323 ; Eldred, 933 F.3d at 119 ; Werdene, 883 F.3d at 217 ; Henderson, 906 F.3d at 1119. We agree. Our view that the agents did not unreasonably exceed the scope of the warrant also comports with the Supreme Court’s standard for demonstrating objective bad faith in executing a warrant. The defendant must show “conscious or flagrant” 10 Nos. 18-2993 & 19-1576 disregard of the warrant’s scope. See Rawlings v. Kentucky, 448 U.S. 98, 110 (1980). The question is whether the officers’ execution of the warrant was “objectively understandable and reasonable.” Maryland v. Garrison, 480 U.S. 79, 88 (1987) (applying good-faith exception to officer’s reliance on warrant with ambiguous scope); see also Messerschmidt, 565 U.S. at 555 (applying similar standard in examining whether ambiguous warrant was facially invalid). “The NIT warrant specifies into which homes an intrusion is permitted (those where the activating computers are located), and on what basis (that the users in those homes logged into Playpen).” Levin, 874 F.3d at 323 . We cannot infer that agents believed themselves to be cabined in the Eastern District of Virginia but flagrantly disregarded that boundary when the stated purpose of the warrant, as issued, was to uncover the unknown locations of anonymous users. Finally, we recently concluded that the Fourth Amendment does not require a tracking warrant to specify a geographic scope at all (even if other laws do). See United States v. Brewer, 915 F.3d 408 , 413–14 (7th Cir. 2019). Under the circumstances, the FBI agents could have reasonably believed that “any” computer used to log into Playpen was within the scope of the warrant, not just those in the Eastern District of Virginia. The district court did not err by denying the motion to suppress. III. Sentencing Grisanti argues that the district court erred procedurally by sentencing him based on improper comparisons to an unrelated defendant and unfounded speculation that he had a medical condition that would inevitably cause him to commit another child-pornography offense. The government responds in three parts: (1) Grisanti waived this argument by Nos. 18-2993 & 19-1576 11 assenting to the judge’s remarks; (2) Grisanti invited any error by asserting that he had an addiction; and (3) the judge did not err by accepting Grisanti’s argument that he had an illness and encouraging him to obtain treatment. We agree with the government that Grisanti invited the judge’s comments about his purported addiction and the need for and possible benefit of treatment. Grisanti asserted just that in his sentencing memorandum and again at the hearing. “A party may not ‘invite’ error and then argue on appeal that the error for which he was responsible entitles him to relief.” United States v. Gaya, 647 F.3d 634, 640 (7th Cir. 2011) (alterations and citations omitted). Grisanti’s choice to focus on an asserted need for professional help was “purposeful; it was part of a strategy” to mitigate his culpability for his crimes. See United States v. Addison, 803 F.3d 916, 920 (7th Cir. 2015). “It is not our job to rescue [Grisanti] from the consequences of that strategic choice.” Id. Grisanti counters that he may have agreed that he had an addiction, but he never suggested that he was likely to commit more crimes. See Zedner v. United States, 547 U.S. 489 , 505–06 (2006) (addressing argument on appeal where party’s contentions were not inconsistent with earlier position). Yet the risk of future crimes seems to be where his addiction argument pointed. Even if Grisanti did not invite the judge’s remark about becoming a repeat offender, there was no error here. The judge did not rely on conjecture, extraneous factors, or her own medical opinion that Grisanti has an addiction that is likely to cause future crimes. In fact, the judge asked whether a psychologist had evaluated him. Rather than suggest that Grisanti would inevitably commit new crimes, the judge expressed optimism that he could succeed and realize 12 Nos. 18-2993 & 19-1576 his potential. She did not find that Grisanti suffers from an illness beyond his control that would render attempts at treatment futile. See United States v. Adams, 646 F.3d 1008, 1012 (7th Cir. 2011). And judges “routinely” make predictions about a defendant’s future conduct, so the judge did not err by reasoning that Grisanti might reoffend if he did not get treatment. See United States v. Kluball, 843 F.3d 716, 720 (7th Cir. 2016). Indeed, a judge must consider any need for “correctional treatment.” 18 U.S.C. § 3553 (a)(2)(D). The judge’s statements also do not show that she equated Grisanti with the other defendant she mentioned, who had just been sentenced for a second offense. In context, the judge was only commenting on her desire that Grisanti obtain treatment and avoid reoffending. Grisanti also asserts that the judge erred by failing to address what he now calls his “principal” mitigation argument, that he deserved a shorter sentence because he had not committed a “contact sex offense.” Grisanti never mentioned this point at the sentencing hearing. Neither the district judge nor we can treat it as central to his plea for leniency. Grisanti advanced this contention in his sentencing memorandum as part of his broader argument that he was unlikely to reoffend, a subject that the judge addressed at length. A judge must “meaningfully” consider the defendant’s principal arguments, but that requirement “does not apply mechanically.” United States v. Hancock, 825 F.3d 340, 343 (7th Cir. 2016), quoting United States v. Estrada-Mederos, 784 F.3d 1086, 1091 (7th Cir. 2015). The district judge discussed the studies that Grisanti cited about recidivism and the reasons he argued that he was unlikely to offend; the explanation was sufficient. Nos. 18-2993 & 19-1576 13 We must also note that Judge Pratt properly rejected Grisanti’s suggestion—based on the Child Pornography Offender Risk Tool and the Correlates of Admission of Sexual Interest in Children assessment—that he is less likely to commit future crimes because he is white. Subject to constitutional limits, sentencing judges have broad discretion about the information they may consider when deciding on an appropriate sentence. See Dean v. United States, 137 S. Ct. 1170, 1175 (2017), citing Pepper v. United States, 562 U.S. 476 , 487–89 (2011); see also 18 U.S.C. § 3661 . But imposing different sentences based on race would violate the Equal Protection Clause of the Fourteenth Amendment and the Sentencing Guidelines. See U.S.S.G. § 5H1.10; United States v. Campbell, 813 F.3d 1016, 1018 (7th Cir. 2016). The Supreme Court in other contexts has permitted consideration of race to offset negative effects of past discrimination and to combat harmful stereotypes. See generally Grutter v. Bollinger, 539 U.S. 306 , 328–30 (2003) (upholding public law school’s affirmative action program). On the other hand, the use of actuarial tools that use race as a factor for assessing probabilities of future crimes, like the studies cited by Grisanti, carry “the potential to reify, rather than ameliorate, extant racial disparities.” See J.C. Oleson, Risk in Sentencing: Constitutionally Suspect Variables & Evidence-Based Sentencing, 64 SMU L. Rev. 1329 , 1387 (2011). The Supreme Court has also recognized that some sentencing differentials “foster[] disrespect for and lack of confi- dence in the criminal justice system because of a widely-held perception that [they] promote[] unwarranted disparity based on race.” Kimbrough v. United States, 552 U.S. 85, 98, 101 (2007) (internal quotation marks omitted) (citing United States Sentencing Commission report and holding that courts may sentence based on policy considerations). The district judge did 14 Nos. 18-2993 & 19-1576 not err by recoiling from Grisanti’s reliance on studies that factor in race. Finally, Grisanti contends that his sentence is substantively unreasonable because the application of U.S.S.G. § 2G2.2, which applies to his child-pornography offenses, resulted in an unreasonably long guideline range. He asserts that some federal courts routinely vary below this guideline on the ground that it is “flawed.” See United States v. Halliday, 672 F.3d 462, 473 (7th Cir. 2012). A sentencing court may reject any guideline on policy grounds as long it acts reasonably in doing so. E.g., United States v. Corner, 598 F.3d 411, 415 (7th Cir. 2010) (en banc). A sentencing court cannot, however, be required to accept a policy argument that rejects a guideline. To sustain the presumption that a within-guideline sentence is reasonable, “a district court need provide only a justification … adequate to allow for meaningful appellate review and to promote the perception of fair sentencing.” United States v. Horton, 770 F.3d 582, 585 (7th Cir. 2014), quoting United States v. Pilon, 734 F.3d 649, 656 (7th Cir. 2013). Here, the judge appropriately discussed Grisanti’s history and characteristics (describing a “deceitful” man who was “not naïve” about his actions), the seriousness of his crimes and the need to protect the public (noting that he had worked near children and the offense involved prepubescent children), and the need to promote respect for the law (observing that he destroyed evidence and implicitly blamed his wife and co-workers). And though the judge commented on Grisanti’s defense, she properly declined to view as an aggravating factor his decision to exercise his right to go to trial. The judgment of the district court is AFFIRMED.