Policy Paper
May 21, 2012

Since September 11, 2001, more than 300 U.S. residents have been prosecuted for crimes related to homegrown terrorism. About half were targeted by law enforcement using infiltration techniques – confidential informants, undercover operations, or, in some cases, both.[i]

The use of infiltration has grown increasingly controversial, particularly within the American Muslim community, where many view these techniques as bordering on entrapment (regardless of the legal definition). In the worst light, informants and undercover officers are seen as agents provocateurs – government employees who are instructed to provoke people into illegal acts so that they can be prosecuted.

A 2011 Pew survey found that 52 percent of all American Muslims feel anti-terrorism policies in the United States single out Muslims for surveillance and monitoring. That number jumped to 71 percent among native-born American Muslims. Forty-four percent of the general public agreed with the assessment.[ii]

That survey was taken before a series of investigative reports by the Associated Press on the New York Police Department’s anti-terrorism unit confirmed some of the community’s worst fears. The AP series, starting in August 2011, revealed that the NYPD has engaged in widespread surveillance of Muslim communities in New York, often without evidence of illegal activity and often without producing actionable results.[iii]

But questions about when and how to use infiltration techniques are not new and are not limited to Muslim communities in the post-9/11 era. Whether it’s an undercover agent buying supplies for an al Qaeda sympathizer,[iv] an FBI handler getting inappropriately close to a mob informant,[v] or informants collecting evidence even as hackers compromise a private company’s data,[vi] infiltration inherently involves unique risks along with its potential rewards.

Informants and undercover agents are essential tools for law enforcement officials, whether they are investigating terrorism, organized crime, corporate wrongdoing, computer hacking, or fraud. Infiltration methods have a proven track record as far as their legality and their investigative merits in the vast majority of cases. But methods that fall within legal bounds are not necessarily without negative consequences. Violent extremism presents a particularly devilish problem set in this respect.

Law enforcement has an obvious stake in trying to determine whether nonviolent people who espouse extremist beliefs or engage in violent rhetoric might become violent. Now more than ever, the government is focused on preventing potential extremists from becoming violent. But infiltration techniques can have a dramatic effect on how targeted communities view efforts to counter and prevent violent extremism by raising questions about the government’s intent and integrity. Aggressive infiltrations can even reinforce extremist narratives that claim the government targets communities because of their fundamental identities rather than in pursuit of illegal activities. The ripple effects of perceived overreach can also make it more difficult for otherwise friendly community partners to encourage cooperation with law enforcement.

Studying the secondary effects of infiltration in current situations verges on the impossible for several reasons. The current activities of undercover agents and informants are closely guarded secrets, for obvious reasons. When infiltrators are exposed in the course of an arrest and prosecution, descriptions of their activities are carefully controlled by prosecutors and carefully spun by defense attorneys, resulting in a distorted picture. Members of a targeted community may be reluctant to frankly discuss their attitudes toward these activities for fear of being targeted themselves. And finally, the secondary effects of infiltration can play out over years or even decades, rendering any short-term picture incomplete.

Legacy cases, while still subject to many of the same pressures, offer an improved window on the details of specific infiltrations and on the medium- and long-term effects on targeted communities. In the case of closed investigations, it is also possible to make a better evaluation of overt successes, such as arrests, prosecutions, and the prevention of violence and other planned activities.

From 1991 to 1993, the FBI conducted an ambitious infiltration program code-named PATCON, short for Patriot Conspiracy. “Patriot” is an umbrella label for a loosely defined movement of antigovernment, racist, anti-Semitic, and/or Christian extremists. The PATCON program is documented in extraordinary detail in thousands of pages of FBI records obtained through the Freedom of Information Act. Interviews with people involved on both sides of the infiltration supplement this information.

PATCON consisted primarily of three FBI undercover agents posing as members of a fictional extremist group called the Veterans Aryan Movement. Three Patriot groups were the primary targets of PATCON – Civilian Materiel Assistance, the Texas Light Infantry, and the American Pistol and Rifle Association. PATCON agents roved the country for more than two years collecting intelligence on these and other Patriot organizations and on dozens of individuals, investigating leads on plots from the planned murder of federal agents to armed raids on nuclear power plants to a new American Revolution.

Despite spending hundreds of thousands of dollars and logging uncounted man-hours, PATCON and related investigations produced negligible results in terms of serious criminal convictions. Instead, PATCON became an intelligence tool, predicated on a series of suspected crimes, most of which were discussed but never committed.

Because it was not directly tied to a prosecution, the existence of PATCON was not formally disclosed until 2007, when references to the program appeared in documents released through Freedom of Information Act requests about the targeted groups. But the Patriot movement was keenly aware it had been infiltrated. Several people involved with targeted groups asserted in interviews after the fact that they were aware that at least one undercover agent involved with PATCON was a “fed.” Whether or not that’s true, the movement was actively worried about infiltration, and justifiably so. One group targeted by PATCON was described by the FBI as “extremely sensitive to investigative pressure.”[vii] Another group deliberately discussed “exotic” threats in order “to provoke the FBI into overreacting and to surface informants.”[viii]

The paranoia that resulted from the awareness of infiltration led to some members being expelled or ostracized on the often-incorrect suspicion that they were informants. Patriot gatherings were at times disrupted and even canceled over concerns about infiltration by federal investigators. One former informant interviewed for this report said he believed the FBI was just fine with that outcome but added that such mind games often reinforced the radical beliefs of those being monitored.

“I think they played a lot of people against each other,” he said. “The guys hated them more and more for it. They thought Big Brother was moving in on them.”

In addition to the murky complexities of the observer effect – in which the act of observing something causes changes in the subject of observation – the PATCON investigation highlights several critical issues surrounding the use of infiltration.

To read the rest of this policy paper, please click here.


[ii]Muslim Americans: No signs of growth in alienation or support for extremism, Pew Research Center, August 2011,

[iii]A list of stories and related documents published by the Associated Press can be found at

[iv]Abby Goodnough, “Man Is Held in a Plan to Bomb Washington,” New York Times, September 28, 2011,

[v]Shelley Murphy, “’Whitey’ Bulger’s ex-girlfriend testifies at Connolly trial,” Boston Globe, October 16, 2008,

[vi]Paul Wagenseil, “FBI may have known in advance of Stratfor hack,”, March 7, 2012,

[vii]FBI communication, San Antonio to HQ, June 16, 1992

[viii]FBI communication, Birmingham to HQ, March 5, 1993