The word "conspiracy" has acquired two strangely contradictory meanings in modern political life. In popular debate, calling something a conspiracy is often a way of dismissing it: the speaker is imagined wearing a tinfoil hat and connecting unrelated dots on a basement wall. In criminal law, however, conspiracy is deadly serious. Once conduct can be characterised as participation in a criminal agreement, speech and association that would ordinarily receive strong constitutional protection can become evidence of a crime.
That is the problem Adam Goldstein explores in a fascinating recent essay for the Foundation for Individual Rights and Expression, FIRE. His argument is built around two men who could scarcely be further apart politically: pro-Trump internet provocateur Douglass Mackey and former CNN presenter Don Lemon. Both found themselves prosecuted under the same federal conspiracy-against-rights statute, 18 U.S.C. ยง241, by governments of opposite political persuasions. Mackey was charged under the Biden administration; Lemon under Trump.
Goldstein's central warning is captured beautifully by his title: "If everything's a conspiracy, nothing's a conspiracy." The point is not that conspiracies do not exist. They obviously do. Organised crime, terrorism, political corruption and ordinary criminal enterprises frequently depend upon people secretly agreeing to commit unlawful acts. The problem arises when "conspiracy" becomes so elastic that agreement is inferred merely because people hold similar opinions, communicate with one another, attend the same events, repeat similar messages or sympathise with the same cause.
The history of the law makes the danger particularly striking. Section 241 ultimately descends from Reconstruction-era legislation enacted against organised violence designed to prevent newly enfranchised black Americans from exercising their constitutional rights. In South Carolina and elsewhere, the Ku Klux Klan used beatings, intimidation and murder as political weapons. The federal government required some means of reaching groups of men who were not merely expressing ugly opinions but coordinating violence to destroy the political rights of their fellow citizens.
That is what a conspiracy looks like in its clearest form. There are multiple people. There is an agreement. There is an unlawful objective. Participants knowingly join the enterprise. Their individual actions may differ, but those actions contribute towards accomplishing the agreed criminal purpose. The distinction between that and merely behaving similarly is fundamental.
Goldstein illustrates it with one of the more entertaining episodes in American organised-crime history. In 1957 police unexpectedly encountered scores of Mafia figures gathering at the rural New York estate of mobster Joseph Barbara. When police arrived, supposedly respectable middle-aged businessmen began fleeing through the surrounding countryside. About 60 organised-crime figures were detained, and many gave remarkably similar explanations for why they happened to be visiting the same gangster on the same afternoon.
It certainly looked suspicious. Prosecutors eventually argued that the similarity of their explanations showed that the mobsters had conspired to obstruct justice and commit perjury. A jury convicted them.
The Second Circuit overturned every conviction. The problem was elementary but enormously important. Similar behaviour did not prove agreement. Each Mafia figure had an obvious individual reason not to tell police what had happened. They could independently decide to be evasive without ever agreeing with one another to lie. As Goldstein neatly summarises the principle, parallel conduct is not conspiracy.
That principle should become considerably more important in the internet age. Millions of people now participate in loosely connected political movements without belonging to anything resembling a traditional organisation. They follow the same accounts, repeat the same slogans, circulate the same memes and descend upon the same targets. A hashtag can generate what appears from the outside to be highly coordinated behaviour even though most participants have never communicated privately with one another.
Ten thousand people can independently decide to share the same post within an hour. To somebody examining the resulting pattern, it may look like an organised campaign. Sometimes it is. Sometimes there is a central group distributing instructions. Sometimes there are private chat groups coordinating participants. But sometimes the explanation is simply that ten thousand people saw something they liked and pressed the same button. The distinction is crucial because otherwise political movements themselves begin to resemble conspiracies.
The Mackey prosecution demonstrates the danger. During the 2016 presidential election, Mackey, operating online as "Ricky Vaughn," distributed deceptive memes telling Hillary Clinton supporters that they could vote by text message. Thousands of people apparently texted the advertised number, although prosecutors produced no evidence that anybody was actually prevented from voting. The memes were reprehensible political dirty tricks, but the government charged Mackey with something more specific: conspiracy to interfere with voting rights.
A jury convicted him, and he received a seven-month prison sentence. Then, in 2025, the Second Circuit unanimously reversed the conviction and ordered an acquittal. The crucial defect was not whether Mackey's memes were admirable, truthful or funny. They plainly were deceptive. The problem was that prosecutors had failed to prove the agreement necessary for conspiracy. Mackey had not even been participating in the private message groups during the period in which the supposed conspiracy was allegedly created.
That is an important distinction in an era when political morality increasingly substitutes for legal analysis. A person can behave badly without committing the particular crime with which prosecutors have charged him. An offensive speaker is not necessarily an inciter. A propagandist is not necessarily a conspirator. A liar is not necessarily part of an organised criminal enterprise.
Don Lemon now presents the same problem from the other side of American politics. In January 2026, anti-ICE protesters entered Cities Church in St Paul, Minnesota, during a Sunday service. They were protesting because one of the church's pastors reportedly worked as an ICE official. Lemon attended with a camera and livestreamed what happened. Prosecutors allege that he did considerably more than observe: the indictment points to his knowledge of the protesters' plans, his withholding of their intended target from his livestream audience, his use of "we" and other conduct allegedly showing participation in the operation. Lemon says he was acting as a journalist covering a newsworthy event.
There is consequently a genuine factual question concerning how close Lemon was to the protest organisation. Journalists do not acquire immunity from ordinary criminal law merely by carrying cameras. A reporter cannot rob a bank alongside three criminals and then invoke press freedom because he was documenting the robbery.
But neither does proximity to wrongdoers transform journalism into conspiracy. A reporter covering extremists may know what they intend to do. He may communicate with them beforehand. He may refrain from revealing their plans because revealing everything would make certain kinds of journalism impossible. He may even sympathise with their political objectives. None of these things necessarily establishes that he agreed to participate in their unlawful conduct.
Goldstein notes that the government's original efforts to obtain arrest warrants for Lemon and his producer encountered judicial resistance. One judge found no probable cause concerning the journalists, and another criticised the government's tendency to lump participants together despite significant differences in what each person had actually done. Prosecutors subsequently obtained a grand-jury indictment, and Lemon is now seeking dismissal. The case therefore remains unresolved.
FIRE's position is particularly interesting because consistency has become so rare in free-speech politics. Mackey is a Right-wing troll. Lemon is a liberal journalist and prominent Trump critic. It would be remarkably easy to decide that Mackey deserved prosecution while Lemon deserves First Amendment protection, or to reverse the formula according to one's politics. But constitutional principles become meaningless if they change whenever the defendant's politics change.
If the government must prove an actual agreement for Mackey, it must prove one for Lemon. If sharing political objectives is insufficient evidence of conspiracy when the defendant is on the Right, it remains insufficient when the defendant is on the Left. If journalists receive protection for reporting on political protesters, the principle cannot depend upon whether the administration approves of the protest.
This is where Goldstein's argument reaches beyond these two prosecutions. Modern politics is increasingly obsessed with networks. Investigators, journalists and activists map relationships between people, organisations, donors, websites and social-media accounts. Person A knows Person B. Person B attended an event organised by Group C. Group C shared an article written by Person D. Person D once appeared on a podcast with Person E. Before long the diagram resembles a bowl of spaghetti and the visual density itself begins to suggest conspiracy. But networks are not agreements.
The same problem afflicts public discussion of "disinformation networks." If thousands of people independently distrust a government policy and repeat similar criticisms, that may indicate coordinated propaganda. It may also indicate that thousands of people independently distrust the policy. Similarity of conclusion cannot itself establish coordination because people exposed to similar evidence frequently reach similar conclusions.
The danger becomes greater when governments enter the picture. Conspiracy law gives prosecutors enormous leverage because ordinary acts can acquire criminal significance when interpreted as steps in a collective enterprise. A meeting becomes evidence. A message becomes evidence. A joke becomes evidence. A social-media repost becomes evidence. Silence can become evidence. Association itself begins to look suspicious.
That is precisely why the agreement requirement matters. It forces the state to prove the bridge between individuals rather than simply pointing to people standing on the same side of the river.
There is a wider cultural lesson here as well. We have become simultaneously too sceptical and too credulous about conspiracies. We ridicule the very idea that powerful people might secretly coordinate to achieve political or economic objectives, even though history contains countless documented conspiracies. At the same time, we increasingly infer coordination whenever people we dislike behave similarly. Both errors arise from abandoning precision.
Watergate involved conspiracy. Organised crime involves conspiracies. Corporate price-fixing can involve conspiracy. Terrorist plots involve conspiracy. Political corruption can involve conspiracy. The category is indispensable precisely because human beings really do secretly cooperate to commit wrongful acts.
But if conspiracy comes to mean nothing more than people sharing objectives, communicating with one another and behaving similarly, the concept ceases to distinguish organised criminal conduct from ordinary social life.
Every political party begins to resemble a conspiracy. Every protest movement does. Every activist network does. Every newsroom coordinating an investigation does. Every group of academics advancing a common intellectual position does. Every online community circulating the same arguments does. That is why Goldstein's title contains more wisdom than first appears. If everything becomes a conspiracy, then the word tells us nothing.
More importantly, the dilution is dangerous to liberty. Criminal law needs distinctions precisely because government possesses powers unavailable to ordinary political combatants. A newspaper columnist can call somebody a conspirator loosely. A prosecutor seeking imprisonment cannot be permitted the same luxury.
The government must prove that particular people actually agreed to pursue a particular unlawful objective. Presence is not enough. Sympathy is not enough. Parallel conduct is not enough. Shared ideology is not enough. Association with unpleasant people is not enough.
That standard will occasionally allow reprehensible people to escape conspiracy convictions. Mackey's deceptive memes do not become admirable because his conviction was overturned. Nor does insisting upon proof in Lemon's case amount to endorsing the disruption of a church service.
It means something more important: criminal liability belongs to acts that the government can prove, not to political identities the government dislikes.
The historical irony is considerable. A statute born from the desperate attempt to destroy the Ku Klux Klan eventually reached a Right-wing internet provocateur posting memes and, under the next administration, a liberal journalist carrying a camera at an anti-ICE protest. The political direction reversed while the temptation remained exactly the same. That should tell us where the enduring danger lies.
Governments change. Prosecutorial powers remain. Every precedent created to punish today's political enemy becomes available for use against tomorrow's. The proper response is therefore not to decide which side deserves the weapon. It is to keep the weapon within carefully defined legal boundaries.
Real conspiracies exist and sometimes cause terrible harm. That is precisely why we need to preserve the meaning of the word. When everything becomes conspiracy, eventually ordinary political association, journalism and dissent can become conspiracy too. And by then the problem is no longer merely that a useful word has lost its meaning. The problem is that an exceptionally powerful criminal concept has lost its limits.
https://www.fire.org/news/if-everythings-conspiracy-nothings-conspiracy