In 1170, England, King Henry II reportedly explained, with respect to Thomas Becket, the Archbishop of Canterbury, “Will no one rid me of this meddlesome priest,” whereupon four Knights travelled to Canterbury and murdered Becket. With modern technology, if Henry had posted a meme under the handle “Curtmantle” with the same exclamation, could he be held criminally liable for the acts of the offending priests? Modern communications technologies permit individuals to reach out to hundreds or millions of anonymous and unidentified individuals, and discuss politics, sports, culture, etc., and act in “concerted actions” with them. Maybe this leads to people dumping ice water on their heads, “surf” on the top of the D train in Queens, or stick metal objects into school-supplied chromebooks (yeah, that’s a thing). Under traditional conspiracy law, if two people agree to perform a criminal act (or a legal act using a criminal means), they each can be prosecuted for the separate crime of “conspiracy” — for the agreement itself. But on the internet, where an anonymous culture of concerted actions exists, how do you apply conspiracy law?

The Meme That Became a Federal Indictment

In United States v. Mackey, No. 23-7577 (2d Cir. July 9, 2025), the U.S. Court of Appeals for the Second Circuit addressed one of the most consequential and novel legal issues of the internet age: Can a person be convicted of a federal conspiracy solely by posting misleading political memes online, without ever having communicated or knowingly coordinated with their alleged co-conspirators?

Douglass Mackey—known by his online alias “Ricky Vaughn”(wild thing… you make my heart sing) —was charged and convicted under 18 U.S.C. § 241 for allegedly conspiring to deprive others of the right to vote by tweeting memes that suggested Hillary Clinton supporters could vote in the 2016 election via text message. The conviction rested not on any direct communication or formal agreement between Mackey and others, but on circumstantial inference from similarity of online content, ideological alignment and his membership in various message boards. The crime was “conspiracy” – the agreement, not actually depriving anyone of their own vote.

The Second Circuit reversed the conviction, emphasizing that while conspiracy law allows for broad attribution of liability, it still requires evidence of a knowing agreement. The mere act of posting or reposting politically motivated messages online, even when false or misleading, does not itself prove a conspiratorial agreement. The court held that the government failed to meet its burden and directed the entry of an acquittal.

The case underscores a critical legal and constitutional dilemma: Conspiracy law has long been known as “the prosecutor’s friend” because of its ability to attribute criminal liability not just for a defendant’s own actions, but for those of others whom they may have never met. This expansive doctrine becomes especially problematic when the underlying acts are themselves expressive conduct, implicating core First Amendment protections.

Federal conspiracy law under 18 U.S.C. § 371 and § 241 is exceptionally potent. Under § 371:
“If two or more persons conspire either to commit any offense against the United States, or to defraud the United States… and one or more of such persons do any act to effect the object of the conspiracy, each shall be fined… or imprisoned…”

Under § 241, the language is similarly broad:
“If two or more persons conspire to injure, oppress, threaten, or intimidate any person… in the free exercise or enjoyment of any right or privilege secured to him by the Constitution or laws of the United States…”

At its core, conspiracy is an inchoate crime. It punishes the agreement itself, regardless of whether the underlying offense is ever completed. As the Supreme Court has long held, the “gist of the crime of conspiracy is the agreement or confederation to do an unlawful act.” United States v. Rabinowich, 238 U.S. 78, 86 (1915).
What makes conspiracy so powerful—and perilous—is its doctrine of attribution. Once an individual knowingly joins a conspiracy, they become liable for all foreseeable acts committed by their co-conspirators in furtherance of the conspiracy, regardless of whether they knew those individuals or agreed to the specific acts. See Pinkerton v. United States, 328 U.S. 640, 646–47 (1946) (“[A]n overt act of one partner may be the act of all without any new agreement specifically directed to that act.”). As the Second Circuit reiterated in United States v. Anderson, 747 F.3d 51, 60 (2d Cir. 2014), a defendant may be convicted so long as they “knowingly joined and participated” in a collective venture with unlawful goals.

The liability persists even if the person later disengages from the conspiracy. To terminate responsibility for future acts, the law requires a formal withdrawal. As the Supreme Court stated in Smith v. United States, 568 U.S. 106 (2013):
“A conspirator must take affirmative steps to disavow or defeat the purpose of the conspiracy… either by informing the authorities or by communicating his withdrawal to his co-conspirators.”

Merely ceasing participation is not enough. In other words, once you’re in the conspiracy, you own the entire enterprise—and unless you actively withdraw, you’re responsible for what your co-conspirators do in your name.

Online “Agreement”

The Mackey case reveals the limits and risks of applying traditional conspiracy doctrine in the decentralized, pseudonymous ecosystem of the internet.
Mackey posted two memes and retweeted another, falsely stating that voters could cast ballots by texting the word “Hillary” to a short code. While the messages were misleading, the government conceded that if Mackey had acted alone, his conduct would not have violated § 241. The crime, according to the prosecution, was not the meme itself, but the supposed agreement behind it.

To prove a conspiracy, the government pointed to Mackey’s prior participation in private Twitter message groups (the “War Room” and others), where other users had discussed spreading similar memes. However, there was no evidence that Mackey participated in or even saw the relevant conversations. He was no longer a member of the chat groups during the key planning period and downloaded the memes from 4chan, not from his alleged co-conspirators.

As the court explained:
“The government was obligated to show that Mackey knowingly entered into an agreement with other people to pursue that objective. This the government failed to do.” Mackey, slip op. at 5.

The government’s fallback theory—that Mackey’s meme-sharing reflected a “silent agreement” to participate in a meme-based voter suppression campaign—was rejected as speculation. Mere ideological similarity, the court emphasized, does not constitute a criminal conspiracy:
“His mere association with those implicated in an unlawful undertaking is not enough to prove knowing involvement.” Id., citing United States v. Nusraty, 867 F.2d 759, 764 (2d Cir. 1989).

The Prosecutor’s Friend Meets the First Amendment

Justice Jackson (Robert H., not Katanji Brown) famously called conspiracy “the darling of the modern prosecutor’s nursery” in Krulewitch v. United States, 336 U.S. 440, 449 (1949) (Jackson, J., concurring), because it allows the state to charge one actor with the deeds of many. This is especially dangerous where the underlying conduct—speech, memes, or political satire—is presumptively protected by the First Amendment. In online spaces, users post, repost and remix content without necessarily knowing where it originated. They may adopt memes ironically or satirically. Yet if the government can attribute criminal purpose based solely on pattern, theme, or ideological alignment, then nearly any participant in a digital forum could be swept into a prosecutable “conspiracy.” The Second Circuit firmly rejected this approach:
“The government must establish that any purpose the defendant and his alleged co-conspirators shared (i.e., knowingly agreed to) was unlawful. It cannot skirt this obligation by framing the purpose at a higher level of generality.” Mackey, slip op. at 36.

Without a clearly proven agreement, criminal liability cannot be based on mere expressive alignment. The internet changes everything—and nothing. It facilitates coordinated speech and disaggregated action. But it does not obviate the fundamental requirement of agreement in criminal conspiracy. The principle that co-conspirators are responsible for each other’s acts—even without direct contact—is troubling enough. But when the underlying conduct is protected speech, the danger is magnified. To avoid criminalizing free expression, courts must vigilantly separate association from agreement and speech from crime.

The Second Circuit’s reversal in Mackey reinforces this constitutional boundary. It affirms that posting online—without more—is not a conspiracy. And it serves as a reminder that the power of conspiracy law must be counterbalanced by the rule of law, due process and the First Amendment.

United States v. Mackey is not merely about election interference or even internet speech. It is about the boundaries of collective liability in a digital world where people coordinate without agreeing, repost without conspiring and speak without acting.

The Second Circuit’s decision reaffirms that conspiracy requires an agreement. It affirms that once joined, a conspiracy imputes sweeping liability—but only if knowingly entered. And it emphasizes that even after joining, responsibility continues until an actor affirmatively withdraws.

In an age where law enforcement is tempted to treat virality as mens rea, Mackey insists on more. It is a cautionary tale against overreach and a necessary reminder that the prosecutor’s friend must be a stranger to the First Amendment.