No. 24-6046

Appeal from the United States District Court for the Western District of Oklahoma (D.C. No. 5:22-CV-00521-G)

Devraat Awasthi, American Civil Liberties Union of Oklahoma Foundation, Oklahoma City, Oklahoma (Megan Lambert of American Civil Liberties Union of Oklahoma Foundation, Oklahoma City, Oklahoma and Jared K. Carter, Cornell Law School First Amendment Clinic, Ithaca, New York, with him on the briefs), for Plaintiffs-Appellants.

Cullen D. Sweeney, Assistant Solicitor General (Garry M. Gaskins, II, Solicitor General and Zach West, Director of Special Litigation, with him on the brief), Office of Attorney General, Oklahoma City, Oklahoma, for DefendantsAppellees.

Page 2Before MATHESON , McHUGH , and ROSSMAN , Circuit Judges.

ROSSMAN , Circuit Judge.

In June 2020, a group of activists were prosecuted in Oklahoma state court for inciting a riot in violation of Oklahoma law. Two years later, those activists (Plaintiffs) sued the Oklahoma Attorney General and Oklahoma County District Attorney (Defendants) in their official capacities under 42 U.S.C. § 1983. Their lawsuit challenged Oklahoma Statutes title 21, section 1311 (Riot Statute) on its face as unconstitutionally overbroad under the First Amendment and unconstitutionally vague under the Fourteenth Amendment. The district court denied Plaintiffs’ request for a preliminary injunction. This appeal followed. Facing issues of first impression involving Oklahoma criminal law, we certified questions to the Oklahoma Court of Criminal Appeals (OCCA).¹ This court expresses its appreciation for the OCCA’s willingness to grant our certification requests.

Page 3Exercising jurisdiction under 28 U.S.C. § 1291, we reject Plaintiffs’ facial challenges to the Riot Statute and affirm.

Plaintiffs are young people from Oklahoma City engaged in local activism. In the spring and summer of 2020, following the police killing of George Floyd in Minnesota, Plaintiffs participated in nightly racial-justice protests outside the headquarters of the Oklahoma City Police Department (OCPD) in downtown Oklahoma City. During those protests, Plaintiffs and others began planning a mural outside OCPD headquarters. The mural, designed by a local artist, would depict “a series of flags honoring Black Lives and symbolizing solidarity, community, and shared struggles, including the Black Liberation Flag, Native American Flag, and the Rainbow Pride Flag.” RI.23–24. On June 22, 2020, a protester obtained a permit to paint the mural, and city employees set up traffic barricades to block off space for painters to work.

Page 4Around 2:30 p.m. on June 23, an OCPD officer moved a traffic barricade near the muralists to access the area with his vehicle. Plaintiffs approached his patrol car and yelled, 'Fuck the police!', 'We have a permit!”, “This is a city ordinance!”, and “Hit me if you want to!” RI.25, 59. The officer reversed his patrol car, turned around, and drove away. Plaintiffs then “ran after the cruiser for a few seconds but did not impede its path out.” RI.25. The muralists continued painting without incident.

A few days later, on June 26, prosecutors charged Plaintiffs in Oklahoma County court with felony incitement to riot, in violation of Oklahoma Statutes title 21, § 1320.2.³ The charging information alleged Plaintiffs, “acting together,” had “surrounded [a] . . . marked patrol vehicle” driven by an officer transporting a homicide witness to OCPD headquarters “and confined [the officer] for a period of time while the defendants yelled threats and expletives[.]' 4 RI.86. Plaintiffs Sincere Terry, Mia Hogsett, Page 5Tyreke Baker, Preston Nabors, and Trevour Webb eventually pleaded guilty to misdemeanor charges.⁵

On June 23, 2022, Plaintiffs brought a lawsuit under 42 U.S.C. § 1983 in federal district court against the Oklahoma Attorney General and the Oklahoma County District Attorney in their official capacities. Their complaint facially challenged the constitutionality of the Riot Statute, OKLA. STAT. tit. 21, § 1311, and sought declaratory and injunctive relief. The Riot Statute states: “Any use of force or violence, or any threat to use force or violence if accompanied by immediate power of execution, by three Page 6or more persons acting together and without authority of law, is riot.” OKLA. STAT. tit. 21, § 1311.

The complaint stated two counts. In count one, Plaintiffs claimed the Riot Statute is unconstitutionally overbroad in violation of the First Amendment. See RI.41 (citing Broadrick v. Oklahoma , 413 U.S. 601, 615 (1973)). In count two, Plaintiffs claimed the Riot Statute is unconstitutionally vague in violation of the Due Process Clause of the Fourteenth Amendment.

Plaintiffs also moved for a preliminary injunction against the enforcement of the Riot Statute.⁶ They acknowledged their burden to satisfy the four preliminary-injunction factors. This appeal turns on the first factor: whether Plaintiffs have demonstrated “a substantial likelihood that they will ultimately succeed on the merits of their suit[.]” Rocky Mountain Gun Owners v. Polis , 121 F.4th 96, 112 (10th Cir. 2024) (citing Winter v. Nat. Res. Def. Council, Inc. , 555 U.S. 7, 20 (2008)).

Plaintiffs argued they could show a substantial likelihood of success on both facial challenges. As to overbreadth, Plaintiffs began by assertingPage 7 “[n]early all speech is protected by the First Amendment,” but one “narrow exception is true threats.” RI.61 (citing Virginia v. Black , 538 U.S. 343, 344 (2003)). A true threat, Plaintiffs explained, is “a serious expression of an intent to commit an act of unlawful violence to a particular individual or group[.]' RI.61 (quoting Black , 538 U.S. at 344). In Plaintiffs' view, a prosecution for a “true threat” requires the State to “consider the mental state of the speaker” and prove “the speaker intend[ed] to intimidate the target of the threat.” RI.61. Plaintiffs argued the Riot Statute violated this requirement, because the law does not “include an intent element.” RI.64. With no mens rea requirement in the statute, “Section 21-1311 criminalizes any threat'-'includ[ing] protected expressive activity such as political hyperbole'-and not just 'true' threats uttered by someone who in fact intended to threaten. RI.65. 'As a result of the statute's overbreadth,' Plaintiffs said, 'speakers at demonstrations have and will restrict their speech, avoiding even constitutionally protected political hyperbole, for fear of prosecution under Section 21-1311.” RI.68 n.5.

As to vagueness, Plaintiffs explained 'void-for-vagueness doctrine requires that a penal statute define the criminal offense with sufficient definiteness that ordinary people can understand what conduct is prohibited and in a manner that does not encourage arbitrary and discriminatory enforcement.” RI.69 (quoting United States v. Hunter, 663Page 8 F.3d 1136, 1141 (10th Cir. 2011)). Plaintiffs argued the Riot Statute “must be found void for vagueness' because it 'does not include a mens rea requirement for a person to be held criminally liable for participating in a riot.” RI.70. The lack of a mens rea requirement “means the statute does not provide clear notice to the ordinary Oklahoma citizen as to the difference between standing peacefully on the street in a protest that happens to turn violent due to other malicious actors and actually being the actor who throws a Molotov cocktail or smashes a car.” RI.70. Defendants opposed the motion, arguing Plaintiffs are unlikely to succeed on the merits.

The district court denied the preliminary injunction request in a written order. Because Plaintiffs sought “a disfavored injunction” against “enforcement of a state law that has been on the books for over 100 years[,]” the district court ruled Plaintiffs faced a 'heightened standard' on their motion, requiring 'a strong showing' on the likelihood-of-success factor. RIII.312–13 (quoting Mrs. Fields Franchising, LLC v. MFGPC , 941 F.3d 1221, 1232 (10th Cir. 2019)). The district court then “construe[d]” the Riot Statute, observing “it is impossible to determine whether a statute reaches too far without first knowing what the statute covers.” RIII.313 (quoting United States v. Williams , 553 U.S. 285, 293 (2008)). The court acknowledged 'state courts are the final arbiters of state law.' RIII.314Page 9 (quoting United States v. DeGasso , 369 F.3d 1139, 1145 (10th Cir. 2004)). But when no controlling state law exists, the district court explained, “the federal court must attempt to predict what the state’s highest court would do.' RIII.314 (quoting DeGasso , 369 F.3d at 1145). Mindful of these principles, the district court recognized it was “without power to adopt a narrowing construction of a state statute unless such a construction is reasonable and readily apparent.” RIII.314 (quoting Stenberg v. Carhart , 530 U.S. 914, 944 (2000)).

The district court discerned a 'reasonable and readily apparent' construction of the Riot Statute in Oklahoma law. Relying on Crawford v. Ferguson , the district court observed the OCCA, more than a century earlier, held that “to be guilty of riot, three or more persons acting together must 'seek to accomplish any unlawful purpose.'' RIII.316 (quoting Crawford v. Ferguson , 115 P. 278, 279 (Okla. Crim. App. 1911)). Then relying on Casteel v. State , the district court recognized the OCCA required the charging document for riot under § 1311’s predecessor statute to allege defendants “acted together willfully and with a common intent to use force or violence.” RIII.317 (citing Casteel v. State , 161 P. 330, 330 (Okla. Crim. App. 1916)). Accordingly, the district court concluded the OCCA 'has interpreted section 1311 to require elements regarding intent that are not Page 10expressly stated in the statute.” RIII.316. Putting those elements together, the district court arrived at the following construction of the Riot Statute:

Any use of force or violence, or any threat to use force or violence if accompanied by immediate power of execution, by three or more persons acting together, willfully , without authority of law, and sharing a common intent to use force or violence or to unlawfully threaten to use force or violence .

RIII.318 (emphasis added to reflect the district court’s narrowing language not originally found in the text of the statute).

In the district court's view, its construction of the Riot Statute comported with Oklahoma law and with Counterman v. Colorado , 600 U.S. 66 (2023), which issued after the parties had briefed the preliminaryinjunction motion.⁷ In Counterman , the Supreme Court held 'the State must prove in true-threats cases that the defendant had some understanding of his statements’ threatening character[,]” and the “precise mens rea standard” sufficient for First Amendment purposes is recklessness. 600 U.S. at 73. “A person acts recklessly, in the most common formulation, when he ‘consciously disregard[s] a substantial [and unjustifiable] risk that the conduct will cause harm to another.’” Id. at 79 (alterations in original) (quoting Voisine v. United States , 579 U.S. 686, 691Page 11 (2016)); see id. (holding, in the language of true threats, recklessness “means that a speaker is aware that others could regard his statements as threatening violence and delivers them anyway” (internal quotation marks omitted)).

The district court found Plaintiffs failed to make a “strong showing of a likelihood of success on the merits” on either of their facial constitutional challenges. RIII.325. As to overbreadth, the court concluded Plaintiffs had failed to show “section 1311 criminalizes a substantial amount of protected speech.' RIII.321. Although Plaintiffs had offered various hypotheticals ostensibly showing the Riot Statute's overbreadth, the district court concluded none 'would constitute actionable riot under section 1311.' RIII.319. As to vagueness, the district court rejected Plaintiffs’ argument that the Riot Statute lacked a mens rea . As construed by the district court, the statute's mens rea of 'willfully' and its three-or-more-people and common-intent elements limit the scope of possible prosecutions and “provide sufficient notice to peaceful protesters as to what conduct constitutes riot under section 1311.” RIII.323. Because Plaintiffs could not meet the likelihood-of-success factor, the district court declined to address the other preliminary-injunction factors and denied the motion.

Page 12This timely appeal followed.⁸

Plaintiffs ask us to reverse the district court's order denying their motion for a preliminary injunction. 'A preliminary injunction is an extraordinary remedy, the exception rather than the rule.” United States ex rel. Citizen Band Potawatomi Indian Tribe of Okla. v. Enter. Mgmt. Consultants, Inc. , 883 F.2d 886, 888 (10th Cir. 1989). To prevail on a preliminary injunction motion, the moving party must prove “(1) that she’s substantially likely to succeed on the merits, (2) that she’ll suffer irreparable injury if the court denies the injunction, (3) that her threatened injury (without the injunction) outweighs the opposing party's under the injunction, and (4) that the injunction isn’t adverse to the public interest.” Free the Nipple-Fort Collins v. City of Fort Collins , 916 F.3d 792, 797 (10th Cir. 2019) (internal quotation marks omitted).

Page 13'An injunction can issue only if each factor is established.' Denver Homeless Out Loud v. Denver , 32 F.4th 1259, 1277 (10th Cir. 2022) (citing Winter , 555 U.S. at 23-24). Where, as here, 'a party seeks a preliminary injunction on the basis of a potential constitutional violation, the likelihood of success on the merits often will be the determinative factor.” Obama for Am. v. Husted , 697 F.3d 423, 436 (6th Cir. 2012) (internal quotation marks omitted); see also Hobby Lobby Stores, Inc. v. Sebelius , 723 F.3d 1114, 1145 (10th Cir. 2013) (noting that “in First Amendment cases, the likelihood of success on the merits will often be the determinative factor' (quoting ACLU of Ill. v. Alvarez , 679 F.3d 583, 589 (7th Cir. 2012))). If a requested injunction would 'change[] the status quo,' the preliminary injunction motion is “disfavored,” and “the moving party faces a heavier burden on the likelihood-of-success-on-the-merits and the balance-of-harms factors[.]' Free the Nipple-Fort Collins , 916 F.3d at 797. To prevail, the movant thus 'must make a strong showing that these [factors] tilt in her favor.' Id. (internal quotation marks omitted). The parties agree Plaintiffs seek a disfavored injunction and must satisfy a “heavier burden[.]” Id.

'District courts have discretion over whether to grant preliminary injunctions, and we will disturb their decisions only if they abuse that discretion.” Courthouse News Serv. v. N.M. Admin. Off. of Cts. , 53 F.4th 1245, 1254 (10th Cir. 2022) (internal quotation marks omitted). “A district Page 14court's decision crosses the abuse-of-discretion line if it rests on an erroneous legal conclusion or lacks a rational basis in the record.' Id. (internal quotation marks omitted). “In reviewing a district court’s decision to grant or deny a preliminary injunction, we thus examine the court's factual findings for clear error and its legal conclusions de novo.' Id. at 1254–55 (internal quotation marks omitted).

On appeal, Plaintiffs insist the district court erroneously concluded they failed to show a substantial likelihood of success on the merits of their facial constitutional challenges to the Riot Statute. This appeal, therefore, presents two issues. First , is the Riot Statute unconstitutionally overbroad under the First Amendment? Second , is the Riot Statute unconstitutionally vague under the Fourteenth Amendment? In light of the parties’ briefing, the record on appeal, and oral argument, the resolution of these issues narrowed further: What mens rea —if any—does the Riot Statute require? We determined that neither the statute’s text nor existing Oklahoma law settled this question.

We then twice certified questions to the OCCA. Granting our requests, the OCCA twice provided answers about Oklahoma criminal law. See Terry v. Drummond ( Terry I ), 2025 OK CR 11, ¶¶ 1-2 (Okla. Crim. App. 2025) (answering initial question); Terry v. Drummond ( Terry II ), 2026 OK CR 10, ¶ 18 (Okla. Crim. App. 2026) (answering second set of questions). The Page 15certification proceedings, and the OCCA’s two opinions issued in response to the certification requests, figure prominently in this appeal. So do the parties’ arguments about Counterman , 600 U.S. 66 (2023).

On March 3, 2025, we certified the following question to the OCCA: 'Does . . . § 1311 require the State to prove the defendant consciously disregarded a substantial risk that his communications would be viewed as threatening violence toward another to convict for threats constituting riot?” Terry v. Drummond , 2025 WL 707451, at *10 (10th Cir. Mar. 3, 2025) (unpublished).

On July 17, 2025, the OCCA unanimously answered the certified question 'in the negative' and stated its reasoning in a short opinion. Terry I , 2025 OK CR 11, ¶¶ 2, 8. The OCCA summed up its 'hold[ing]': '[T]he State in a charge of riot must prove the defendant's mutual or common intent with two or more others to use or threaten violence, accompanied by an immediate power to carry the threat into being.” Id. at ¶ 9. The OCCA determined the Riot Statute lacks the Counterman -required recklessness element but insisted that omission creates no constitutional problem. Endorsing the district court’s understanding of the Riot Statute, the OCCA found “the combined factual elements of a mutual or common intent to use or threaten violence, and that the violent threats bePage 16 ‘accompanied by immediate power of execution,’ substantially produce the same effect' as the reckless-subjective-intent element announced in Counterman . Id. at ¶ 8 (quoting OKLA. STAT. tit. 21, § 1311). “In this way,” the OCCA reasoned, “the statute precludes punishment for mere hyperbole, idle threats, angry bluster, or lawful protest.' Id. at ¶ 9. Notably, the OCCA did not mention 'recklessness,' subjective intent, or Counterman . Its 'hold[ing]' also did not speak in terms of 'willfulness.'⁹

On October 24, 2025, we certified more questions to the OCCA. We undertook this additional request 'cautiously, careful not to 'trouble our sister state courts every time an arguably unsettled question of state law comes across our desks.’” Terry v. Drummond , 2025 WL 3012448, at *1 (10th Cir. Oct. 24, 2025) (unpublished) (quoting Pino v. United States , 507 F.3d 1233, 1236 (10th Cir. 2007)). But after “careful consideration” of Terry I , “additional questions of state law—both identified in the original Page 17certification order and key to the resolution of this appeal-remain outstanding.” Id. Specifically, we sought to understand the meaning of 'willfulness' under Oklahoma criminal law.¹⁰ See id.

Page 18On March 5, 2026, the OCCA answered our additional questions in another opinion.¹¹ See Terry II , 2026 OK CR 10, ¶ 18. The OCCA “explicitly adopt[ed] the definition of willfulness' in Oklahoma Statutes title 21, section 92 “as the willfulness required under state law for a threat-based violation of” the Riot Statute. Terry II , 2026 OK CR 10, at ¶ 7. That is, the OCCA made clear a threat-based prosecution under the Riot Statute requires “a purpose or willingness to commit the act or the omission referred to. It does not require any intent to violate law, or to injure another, or to acquire any advantage.” Id. (quoting OKLA. STAT. tit. 21, § 92). The OCCA elaborated that “willfulness” under § 92 means “‘intentionally,’ ‘designedly,’ 'without lawful excuse'-that is, not accidentally.' Id. at ¶ 15 (quoting Miller v. State , 130 P. 813, 815 (Okla. Crim. App. 1913)). In the OCCA's view, the mens rea of willfulness, combined with the other elements of the Riot Statute, means a defendant convicted of riot “has at least acted with the type of recklessness required by Counterman [.]” Id. at ¶ 17 (emphasis added).¹²

Page 19With the benefit of the OCCA’s interpretation of Oklahoma criminal law, we turn now to the federal constitutional issues and arguments before us and explain why affirmance is required.¹³

“We review challenges to the constitutionality of a statute de novo.” United States v. Harrison , 153 F.4th 998, 1009 (10th Cir. 2025) (internal quotation marks omitted). We start with overbreadth, then turn to vagueness.¹⁴

“The First Amendment, applicable to the States through the Fourteenth Amendment, prohibits the enactment of laws 'abridging the freedom of speech.'' Reed v. Town of Gilbert , 576 U.S. 155, 163 (2015) (quoting U.S. Const. amend. I). “It has long been recognized that the First Amendment needs breathing space and that statutes attempting to restrict or burden the exercise of First Amendment rights must be narrowly drawn[.]” Broadrick , 413 U.S. at 611. “Overbroad laws ‘may deter or “chill” constitutionally protected speech,’ and if would-be speakers remain silent, society will lose their contributions to the ‘marketplace of ideas.’” United States v. Hansen , 599 U.S. 762, 769–70 (2023) (quoting Virginia v. Hicks , 539 U.S. 113, 119 (2003)).

'To guard against those harms, the overbreadth doctrine allows a litigant (even an undeserving one) to vindicate the rights of the silenced, as well as society’s broader interest in hearing them speak.” Id. at 770. “[T]he overbreadth doctrine instructs a court to hold a statute facially unconstitutional even though it has lawful applications, and even at the behest of someone to whom the statute can be lawfully applied.” Id. at 769. “Invalidation for overbreadth is strong medicine that is not to be casually employed.” Williams , 553 U.S. at 293 (internal quotation marks omitted);Page 21 see also Hansen , 599 U.S. at 769 (calling an overbreadth challenge “unusual”).

“The overbreadth analysis has two steps.” United States v. Streett , 83 F.4th 842, 852 (10th Cir. 2023). “The first step ‘is to construe the challenged statute' to determine whether it covers protected speech.' Id. (quoting Williams , 553 U.S. at 293). This first step proceeds from the common-sense premise that a court must “determine what [the law] covers” before it “can do anything else with these facial challenges[.]” Moody v. NetChoice, LLC , 603 U.S. 707, 725 (2024) (first alteration in original) (quoting Hansen , 599 U.S. at 770).

The second step is 'to determine whether the unconstitutional applications of the statute are 'substantially disproportionate to the statute’s lawful sweep.’” Streett , 83 F.4th at 852 (quoting Hansen , 599 U.S. at 770). “For the statute to be unconstitutionally overbroad, its ‘unconstitutional applications must be realistic, not fanciful[.]’” Id. at 853 (quoting Hansen , 599 U.S. at 770); see also Members of City Council of L.A. v. Taxpayers for Vincent , 466 U.S. 789, 801 (1984) ('[T]here must be a realistic danger that the statute itself will significantly compromise recognized First Amendment protections of parties not before the Court for it to be facially challenged on overbreadth grounds.”). A “lopsided ratio” of unconstitutional applications to constitutional ones suggests impermissible Page 22overbreadth. Hansen , 599 U.S. at 770. The upshot is, in this 'singular context, even a law with ‘a plainly legitimate sweep’ may be struck down in its entirety. But that is so only if the law's unconstitutional applications substantially outweigh its constitutional ones.” Moody , 603 U.S. at 723–24 (emphasis added). “The overbreadth claimant bears the burden of demonstrating, 'from the text of [the law] and from actual fact,' that substantial overbreadth exists.' Hicks , 539 U.S. at 122 (alteration in original) (quoting N.Y. State Club Ass’n, Inc. v. City of New York , 487 U.S. 1, 14 (1988)).

Plaintiffs contend the Riot Statute, on its face, is unconstitutionally overbroad under the First Amendment. We must disagree.

The first step of the overbreadth inquiry requires us to construe the Riot Statute. See Streett , 83 F.4th at 852. Before the OCCA answered our certified questions, it was unclear whether the Riot Statute had any intent requirement. Now, it is undisputed that, under Oklahoma law, the Riot Statute requires the State to prove 'the defendant's mutual or common intent [of willfulness] with two or more others to use or threaten violence, accompanied by an immediate power to carry the threat into being.” Terry I , 2025 OK CR 11, ¶ 9; see also Terry II , 2026 OK CR 10, ¶ 18 (clarifying the mens rea under § 1311 is 'willfully'); Aplt. Second Supp. Br. at 11-14Page 23 (discussing Terry II ’s construction of the Riot Statute); Aple. Second Supp. Br. at 1 (similar). “As Oklahoma’s court of last resort for criminal appeals, the OCCA's interpretation of state law is controlling.' United States v. Faulkner , 950 F.3d 670, 677 n.8 (10th Cir. 2019); accord Hawkins v. Mullin , 291 F.3d 658, 662 (10th Cir. 2002) (recognizing “this court is bound by the [OCCA]’s interpretation of its own law”); see Johnson v. Fankell , 520 U.S. 911, 916 (1997) ('Neither [the Supreme] Court nor any other federal tribunal has any authority to place a construction on a state statute different from the one rendered by the highest court of the State.”).

At the second step of the overbreadth inquiry, Plaintiffs insist that, even as construed by the OCCA, the Riot Statute is 'categorically overbroad.” Aplt. Second Supp. Br. at 3. In Plaintiffs’ view, willfulness “falls far short” of Counterman ’s recklessness standard. Aplt. Second Supp. Br. at 12. Plaintiffs assert willfulness 'merely requires that the defendant voluntarily speaks[,]” Aplt. Second Supp. Br. at 13, whereas the Counterman standard requires the defendant 'consciously disregarded a substantial risk that his communications would be viewed as threatening violence[,]” Counterman , 600 U.S. at 69. That the Riot Statute’s mens rea fails to satisfy the Counterman recklessness standard, Plaintiffs say, reveals its impermissible overbreadth: The statute “lack[s] a constitutionally required element' and so is 'invalid in all applications.'Page 24 Aplt. First Supp. Br. at 4. In the language of overbreadth doctrine, the Riot Statute has no “plainly legitimate sweep.”

To support their overbreadth argument, Plaintiffs offer two hypotheticals. First, Plaintiffs imagine a football game between the University of Mississippi and the University of Oklahoma. If fans chant, “we’re gonna beat the hell out of you!”, Plaintiffs claim those fans “could be prosecuted for a felony under Section 1311 for participating in protected speech.” Op. Br. at 26–27. Second, Plaintiffs conjure a group of protesters who agree to chant 'fight back' in a demonstration against government overreach. Plaintiffs maintain the protesters could be prosecuted under the Riot Statute due to their “immediate power to execute on their threat to 'fight back.'' Aplt. First Supp. Br. at 12. According to Plaintiffs, these hypotheticals reveal the Riot Statute’s overbreadth: “Section 1311 plainly reaches a substantial amount of protected expression.” Op. Br. at 27; Aplt. First Supp. Br. at 12–13.

Plaintiffs have not carried their burden to establish overbreadth. At the second step of the inquiry, a challenger must show a law’s unconstitutional applications are ' substantially disproportionate to the statute's lawful sweep.' Hansen , 599 U.S. at 770 (emphasis added). This requires a court to assess a law’s unconstitutional uses and compare them to its “valid reach.” Id. at 782. But Plaintiffs offer no plausible examples of Page 25the Riot Statute's unconstitutional applications and fail to accurately account for its “lawful sweep.” Id. at 770.

We start by appraising the unconstitutional-applications side of the ledger. Based on Plaintiffs' arguments in this appeal, 'we find it pretty much blank.” Id. at 782. Plaintiffs speculate about prosecutions of football fans taunting their opponents and political protesters yelling generic slogans at demonstrations. But Plaintiffs have not identified any “realworld” examples of improper threats prosecutions in the Riot Statute’s 116year life.¹⁵ United States v. Brune , 767 F.3d 1009, 1021 (10th Cir. 2014)Page 26 (internal quotation marks omitted) (rejecting an overbreadth challenge); see Hansen , 599 U.S. at 782 (faulting the challenger for “fail[ing] to identify a single prosecution for ostensibly protected expression” in the seven decades since Congress enacted the law at issue).

Plaintiffs insist they “need not identify specific historical examples or even particularly credible forecasts of overbroad applications to prevail on their substantial overbreadth claim.” Aplt. First Supp. Br. at 13 n.7. That position misunderstands what the law requires. See Hansen , 599 U.S. at 770 (requiring a First Amendment facial challenge to show unconstitutional applications that are “realistic, not fanciful”); Wash. State Grange v. Wash. State Republican Party , 552 U.S. 442, 450–51 (2008) (instructing courts not to 'go beyond the statute's facial requirements and speculate about ‘hypothetical’ or ‘imaginary’ cases”); Williams , 553 U.S. at 301 (rejecting “an endless stream of fanciful hypotheticals”). Speculation will not satisfy the challenger's burden on overbreadth. See Hicks , 539 U.S. at 122 ('The overbreadth claimant bears the burden of demonstrating, ‘from the text of [the law] and from actual fact ,' that substantial overbreadth exists.' (alteration in original) (emphasis added) (quoting N.Y. State Club Ass’n , 487 U.S. at 14)); e.g. , Streett , 83 F.4th at 855 (finding challenger 'failed to Page 27demonstrate as a matter of 'actual fact' that there [were] a substantial number of . . . potentially unconstitutional applications' of the statute at issue (quoting Hicks , 539 U.S. at 122)); United States v. Ostrander , 114 F.4th 1348, 1362–63 (11th Cir. 2024) (holding an overbreadth challenger 'has not come close to meeting his burden of proving a realistic danger' because, even if the challenger showed “overbreadth from the text of the law,' he still 'has failed to do so in actual fact' (brackets and internal quotation marks omitted)).

To be sure, Plaintiffs suggest their own prosecutions show the overbreadth of the Riot Statute. We acknowledge the facts of an instant case can provide a meaningful data point as to overbreadth. But absent a considerable imbalance between valid and invalid uses, “courts must handle unconstitutional applications as they usually do-case-by-case.' Hansen , 599 U.S. at 770. Based on the allegations and arguments before us, we cannot conclude the record suffices to establish “substantial” overbreadth. Moody , 603 U.S. at 718. Plaintiffs have not shown the Riot Statute has “any realistic danger of chilling the expression of parties not before the court[.]” Faustin v. City & County of Denver , 423 F.3d 1192, 1201 (10th Cir. 2005).

On the “legitimate sweep” side of the ledger, Plaintiffs’ challenge fares no better. Even if Plaintiffs had made a strong showing of plausible constitutional danger, they would also need to show that danger isPage 28 “ substantial , not only in an absolute sense, but also relative to the statute’s plainly legitimate sweep.' Williams , 553 U.S. at 292 (explaining this requirement has been “vigorously enforced” by the Court). “The ‘mere fact that one can conceive of some impermissible applications of a statute is not sufficient to render it susceptible to an overbreadth challenge.’” Id. at 303 (quoting Taxpayers for Vincent , 466 U.S. at 800). Here, Plaintiffs too-readily dismiss this aspect of their burden on overbreadth.

The text of the Riot Statute plainly shows some prosecutions will not involve speech at all.¹⁶ See OKLA. STAT. tit. 21, § 1311 (criminalizing “[a]ny use of force or violence, or any threat to use force or violence if accompanied by immediate power of execution, by three or more persons acting together and without authority of law' (emphasis added)); Okla. Unif. Jury Instructions-Crim. 6-58 (2d ed.) (stating the elements of riot under § 1311 can be shown by '(use of force)/violence . . . OR . . . a threat to use force/violence' (bolding omitted)). Riot prosecutions for constitutionally unprotected conduct are neither hard to imagine nor hard to find. See, e.g. , Schoolcraft v. State , 178 P.2d 641, 643-45, 651 (Okla. Crim. App. 1947)Page 29 (affirming the riot convictions of four individuals who terrorized several locations and violently attacked at least seven victims, some fatally); Perkins v. State , 250 P. 544, 544–45, 548 (Okla. Crim. App. 1926) (affirming riot convictions of Ku Klux Klan members who violently whipped a man as many as thirty times); Darneal v. State , 174 P. 290, 290-91, 293 (Okla. Crim. App. 1917) (affirming a riot conviction of a defendant who in the middle of the night “took an old man” out of the tent where he lived with his two daughters, “dragg[ed] him a short distance,” beat and whipped him, and cut off his hair). Plaintiffs do not attempt to argue such “heartland” prosecutions would run afoul of the Constitution. Hansen , 599 U.S. at 782 (“A brief survey of the Federal Reporter confirms that these are heartland . . . prosecutions.”); see Counterman , 600 U.S. at 73 (limiting its holding to “true-threats cases”).

In sum, based on Plaintiffs' arguments, we cannot say in this case that the law's 'unconstitutional applications substantially outweigh its constitutional ones.” Moody , 603 U.S. at 724 (emphasis added). Measuring de novo the Riot Statute’s unconstitutional applications against its plainly legitimate sweep, we conclude Plaintiffs fail to make a strong showing they are substantially likely to succeed on the merits of their overbreadth claim. See Free the Nipple-Fort Collins , 916 F.3d at 797. To the extent Plaintiffs seek a preliminary injunction based on their claim the Riot Statute is Page 30unconstitutionally overbroad, we affirm the district court’s decision to deny that request. See Denver Homeless Out Loud , 32 F.4th at 1277.

We turn now to vagueness. Arguing the Riot Statute is unconstitutionally vague under the Fourteenth Amendment, Plaintiffs seek reversal of the district court’s decision not to issue a preliminary injunction. Again, reviewing de novo , we discern no error.

'In our constitutional order, a vague law is no law at all.' United States v. Davis , 588 U.S. 445, 447 (2019). “Vague laws contravene the ‘first essential of due process of law’ that statutes must give people ‘of common intelligence' fair notice of what the law demands of them.' Id. at 451 (quoting Connally v. Gen. Constr. Co. , 269 U.S. 385, 391 (1926)). 'Vague laws also undermine the Constitution’s separation of powers” by “hand[ing] responsibility for defining crimes to relatively unaccountable police, prosecutors, and judges, eroding the people’s ability to oversee the creation of the laws they are expected to abide.” Id.

To guard against these harms, 'the void-for-vagueness doctrine requires that a penal statute define the criminal offense with sufficient definiteness that ordinary people can understand what conduct is prohibited and in a manner that does not encourage arbitrary and Page 31discriminatory enforcement.' Kolender v. Lawson , 461 U.S. 352, 357 (1983). Accordingly, a court may invalidate a statute as unconstitutionally vague “for either of two independent reasons. First, if it fails to provide people of ordinary intelligence a reasonable opportunity to understand what conduct it prohibits. Second, if it authorizes or even encourages arbitrary and discriminatory enforcement.' Hill v. Colorado , 530 U.S. 703, 732 (2000). Here, Plaintiffs argue the Riot Statute is vague for both reasons. Where a challenged law “interferes with the right of free speech or of association, a more stringent vagueness test should apply.” Hoffman Estates , 455 U.S. at 499; see also NAACP v. Button , 371 U.S. 415, 432 (1963) ('[S]tandards of permissible statutory vagueness are strict in the area of free expression.”).

Vagueness doctrine does not require 'mathematical certainty' in statutes. United States v. Lesh , 107 F.4th 1239, 1249 (10th Cir. 2024) (quoting Grayned v. City of Rockford , 408 U.S. 104, 110 (1972)). Because the Constitution does not “impose impossible standards of specificity,” courts must 'remain ever mindful that general statements of the law are not inherently incapable of giving fair and clear warning.” Sperry v. McKune , 445 F.3d 1268, 1271 (10th Cir. 2006) (internal quotation marks omitted). “After all, in most English words and phrases there lurk uncertainties, so it is always easy to argue that words are incapable of expressing fixed and determinate concepts.” Wyo. Gun Owners v. Gray , 83 F.4th 1224, 1233–34Page 32 (10th Cir. 2023) (brackets and internal quotation marks omitted). 'What renders a statute vague is not the possibility that it will sometimes be difficult to determine whether the incriminating fact it establishes has been proved; but rather the indeterminacy of precisely what that fact is.' Williams , 553 U.S. at 306.

This appeal began with a dispute about whether the statute has a mens rea element at all. The OCCA has now resolved that question, as all agree: The State must prove in a § 1311 prosecution that the defendant acted “willfully[.]” Terry II , 2026 OK CR 10, ¶ 18. But the consensus about the mens rea in the Riot Statute does not settle Plaintiffs' vagueness challenge just yet. The reason, in Plaintiffs' view, is that the OCCA's construction of the Riot Statute still does not pass federal constitutional muster, so reversal is required. Plaintiffs raise two arguments in support, but neither is persuasive.

First , Plaintiffs argue the mens rea established by Terry II attaches only to a defendant's decision to speak , not to threaten . According to Plaintiffs, the “only mens rea relating to the intent to threaten is ‘common intent.' Neither the statute nor the OCCA's answers give notice of what level of intent to threaten is required.' Aplt. Second Supp. Br. at 14-15. Thus, Plaintiffs say, the OCCA’s opinions “offer no clarity to Oklahomans Page 33seeking to exercise their right to protest while steering clear of potential prosecution under [§] 1311.” Aplt. Second Supp. Br. at 15.

Plaintiffs are incorrect. If a statute has a “general scienter provision,” courts apply that scienter provision to all “statutory terms that ‘separate wrongful from innocent acts.’” Ruan v. United States , 597 U.S. 450, 458 (2022) (quoting Rehaif v. United States , 588 U.S. 225, 229, 232 (2019)); accord MODEL PENAL CODE § 2.02(4) (“When the law defining an offense prescribes the kind of culpability that is sufficient for the commission of an offense, without distinguishing among the material elements thereof, such provision shall apply to all the material elements of the offense, unless a contrary purpose plainly appears.”). This rule has been routinely used to flesh out criminal statutes where one element lacks a scienter provision, but another part of the statute supplies that provision. See, e.g. , United States v. X-Citement Video, Inc. , 513 U.S. 64, 68-72 (1994) (holding the mens rea “knowingly” applied to both subprovisions in a statute, even where that interpretation was not “[t]he most natural grammatical reading” of the statute, because such a reading “avoid[s] substantial constitutional questions”); Ruan , 597 U.S. at 454-55, 458-59 (applying the 'general scienter provision” of “knowingly or intentionally” to the authorization-todistribute-substances element of 21 U.S.C. § 841 because applying the “ mens rea to the authorization clause . . . 'helps advance the purpose of Page 34scienter, for it helps to separate wrongful from innocent acts'' (quoting Rehaif , 588 U.S. at 232)). We readily conclude the 'willfully' mens rea identified by Terry II covers every element required in a threats prosecution under the Riot Statute. Ordinary principles used to interpret criminal statutes undercut Plaintiffs’ contrary argument.

Second , even if “willfulness” applies to every element, Plaintiffs argue that fails to solve the Riot Statute's vagueness problem. The reason, Plaintiffs insist, is the statute’s mens rea of “willfully” does not clear the hurdle of “recklessness” set by the Supreme Court in Counterman . For their part, Defendants say we can reject Plaintiffs’ vagueness challenge merely on the fact that the Riot Statute contains a mens rea of willfulness. In their view, we need not answer in this case the “theoretical question” whether the Riot Statute’s mens rea of willfulness comports with the recklessness standard in Counterman . Aple. Second Supp. Br. at 18. We agree.

As the Supreme Court has repeatedly recognized, “a scienter requirement may mitigate a law’s vagueness, especially with respect to the adequacy of notice to the complainant that his conduct is proscribed.' Hoffman Estates , 455 U.S. at 499 & n.14 (collecting cases); see United States v. El-Hajjaoui , 227 F.3d 1274, 1277 n.1 (10th Cir. 2000) (“Not infrequently the Supreme Court, in passing upon a statute claimed to be unconstitutional for vagueness, has concluded that the statute gives fair Page 35warning because scienter is an element of the offense.” (internal quotation marks omitted)). In other words, “[t]he presence of a scienter inquiry can save an otherwise vague statute.” Ward v. Utah , 398 F.3d 1239, 1252 (10th Cir. 2005) (rejecting a facial vagueness challenge). More particularly, we have found a 'statutory requirement that an act must be willful or purposeful . . . does relieve the statute of the objection that it punishes without warning an offense of which the accused was unaware.' United States v. Corrow , 119 F.3d 796, 804 (10th Cir. 1997) (quoting Screws v. United States , 325 U.S. 91, 102 (1945) (plurality opinion)).

Applying these principles, we reject Plaintiffs’ vagueness challenge. Everyone agrees the OCCA has adopted 'willfully' as the mens rea for threats prosecutions under the Riot Statute. Terry II , 2026 OK CR 10, ¶ 18. The Riot Statute, then, is “not silent as to mens rea [.]” Ruan , 597 U.S. at 458. Under these circumstances, we conclude the Riot Statute’s mens rea requirement “mitigates any vagueness and makes the statute constitutional.'¹⁷ Bushco v. Shurtleff , 729 F.3d 1294, 1307 (10th Cir. 2013) (brackets omitted) (quoting Ward , 398 F.3d at 1252).

Plaintiffs’ request for a preliminary injunction fails the first factor: Plaintiffs have failed to make a strong showing they are substantially likely Page 36to succeed on the merits of their Fourteenth Amendment challenge. See Free the Nipple-Fort Collins , 916 F.3d at 797; Denver Homeless Out Loud , 32 F.4th at 1277 .

We AFFIRM the district court's denial of Plaintiffs' motion for a preliminary injunction.

Page 37No. 24-6046, Terry v. Drummond

ROSSMAN , Circuit Judge, concurring.

As the majority opinion noted, we need not decide in this case whether the willfulness mens rea in the Riot Statute comports with Counterman v. Colorado , 600 U.S. 66 (2023). Though this question is appropriately understood as 'theoretical' in Plaintiffs' appeal, see supra Maj. Op. at 34, I write separately for the day when the question is actually before us.

In Counterman , the Supreme Court considered whether the First Amendment 'requires proof that the defendant had some subjective understanding of the threatening nature of his statements” before a defendant’s threats may be punished by the State. 600 U.S. at 69. Defendant Billy Counterman sent hundreds of Facebook messages to C.W., a local singer and musician whom Mr. Counterman had never met. Some messages were innocuous; others suggested harm.¹ The messages upended C.W.’s daily life, and she contacted law enforcement.

The State of Colorado charged Mr. Counterman under a statute making it unlawful to '[r]epeatedly . . . make[] any form of communication with another person” in “a manner that would cause a reasonable person to suffer Page 38serious emotional distress and does cause that person . . . to suffer serious emotional distress.' Id. at 70 (alterations in original) (quoting COLO. REV. STAT. § 18-3-602(1)(c) (2022)). The only evidence Colorado proposed to introduce at trial were the Facebook messages. Mr. Counterman moved to dismiss the criminal charge on First Amendment grounds, arguing his messages were not so-called “true threats” and were therefore protected by the First Amendment. Id. at 71. Under Colorado law, the prosecution had to show only that a reasonable person would have viewed the messages as threatening—not that Mr. Counterman subjectively understood the threatening nature of the communications. The trial court denied the motion to dismiss, and a jury found Mr. Counterman guilty. The Colorado Court of Appeals affirmed, and the Colorado Supreme Court denied review.

The Supreme Court reversed. The Court reiterated that “[t]rue threats of violence” are among the constitutionally prohibitable categories of speech. Id. at 72. But, the Court recognized, '[p]rohibitions on speech have the potential to chill, or deter, speech outside their boundaries.” Id. at 75. “The speaker’s fear of mistaking whether a statement is a threat; his fear of the legal system getting that judgment wrong; his fear, in any event, of incurring legal costs-all those may lead him to swallow words that are in fact not true threats.” Id. at 78.

Page 39To “prevent that outcome” and “reduc[e] an honest speaker’s fear that he may accidentally . . . incur liability,' id. at 75 (emphasis added) (internal quotation marks omitted), the Supreme Court held a State can punish a threat only by first proving the defendant “had some subjective understanding of the threatening nature of his statements[,]” id. at 69; see id. at 76 (explaining the subjective-intent requirement prevents prosecutions of a 'clueless speaker' whose words were not “‘intended’ . . . to produce imminent disorder” (quoting Hess v. Indiana , 414 U.S. 105, 109 (1973) ( per curiam ))). Specifically, the First Amendment requires a subjective mens rea of at least recklessness, a mental state in which “a speaker is aware that others could regard his statements as threatening violence and delivers them anyway.” Id. at 79 (internal quotation marks omitted).

In my view, the weight of Oklahoma authority, read in light of Terry v. Drummond ( Terry II ), 2026 OK CR 10 (Okla. Crim. App. 2026), settles that the mens rea of willfulness—in the context of a true-threats prosecution under the Riot Statute—is at least equivalent to Counterman 's recklessness standard. The point of the Counterman standard is to prevent the State from punishing speech that threatens inadvertently . Its intended beneficiaries are the “honest speaker[],' 600 U.S. at 75 (internal quotation marks omitted), the 'clueless speaker,” id. at 76, and even the “delusional speaker,” all of whom “may lack awareness of the threatening nature of [their] speech[,]” id. at 120 (Barrett, J., Page 40dissenting). If a speaker is merely negligent—that is, if a speaker “is not but should be aware of a substantial risk” others will understand her speech as a threat—the State cannot prosecute. Id. at 79 n.5.

In Terry II , the OCCA clarified 'willfully' in the Riot Statute means 'intentionally,' 'designedly,' and ' not accidentally .' 2026 OK CR 10, ¶ 15 (emphasis added) (quoting Miller v. State , 130 P. 813, 815 (Okla. Crim. App. 1913)). Put differently, a crime that cannot be committed “accidentally” under Oklahoma law is a crime that requires an intent of at least recklessness. As the Oklahoma Supreme Court recently explained, 'negligence is ordinarily associated with types of inadvertence and not types of willful conduct.” Bailey v. State ex rel. Bd. of Tests for Alcohol & Drug Influence , 2022 OK 50, ¶ 51, 510 P.3d 845, 862 (Okla. 2022); see id. at ¶ 49, 861 ('[O]rdinary and gross negligence . . . both differ in kind from willful and intentional conduct[.]' (quoting Altman v. Aronson , 121 N.E. 505, 592 (Mass. 1919))); Rogers v. Excise Bd. of Greer Cnty. , 701 P.2d 754, 761 (Okla. 1984) (holding “[w]illfulness,” in context of the Oklahoma Open Meeting Act, “encompasses conscious, purposeful violations of the law or blatant or deliberate disregard of the law by those who know, or should know the requirements of the [law]”). Thus, in a prosecution charging true threats, I would conclude the Riot Statute’s mens rea of willfulness satisfies Counterman ’s recklessness requirement.