A framework

Seven forms of weaponization

Only one of them is political.

Say “weaponized” and everyone hears the same thing: a prosecution brought for political reasons. That single assumption is why the word gets waved away — because if weaponization only means political persecution, then proving it means proving what was inside a prosecutor’s head. Nobody can open a head. So the argument stalls in the same place every time, and the person making it walks off sounding like someone making excuses.

Watch how that trap is built. In a federal criminal case the government does not prove what a defendant was thinking either. It cannot. What it does is convince twelve people — from the timing, from what a person did and did not do, from what anyone in that position would have known. Juries are told they may infer intent from circumstance, and that this counts as much as a document.

Turn that same inference around, point it at the prosecution, and it stops being evidence and becomes speculation.

To convict us, what we were thinking can be assumed. To accuse them, what they were thinking must be proven.

So this page asks one question seven times: what was this prosecution for? Politics is one possible answer, and the only one that turns on what somebody intended. The other six answers leave a record — dates, filings, numbers, and the government’s own documents. That is what the rest of this page is made of.

What we are naming

We are done being told that naming this is a cheap excuse — told so by a system that has never once had to explain itself. So here is the definition, and it is narrower than people expect.

Weaponized justice is the machinery of criminal law driven by a purpose other than justice.

We did not get to decide what that purpose is supposed to be. The Supreme Court settled it in 1935, in a sentence quoted in courtrooms ever since: the government’s interest in a criminal prosecution “is not that it shall win a case, but that justice shall be done.”

Berger v. United States, 295 U.S. 78, 88 (1935)

Read that as the standard it is. A prosecution is not an instrument for winning, for advancing, for making an example, for settling a score, or for producing a signature by Friday. When it is used for any of those, it has been weaponized — whether or not anyone involved would ever put it that way, and whether or not politics is anywhere near it.

Which is why we are not going to argue about motive. Purpose is not a state of mind. Purpose is visible in what a prosecution actually did — what it produced, what it cost, what it never bothered to test, and what it stopped caring about the moment it got what it came for. That leaves a record. Motive does not.

Seven purposes. One of them is politics.

The seven forms

Seven answers to the same question: what was served instead?

  1. 01PoliticalCharged for who you are aligned with
  2. 02CareerThe conviction is worth something to someone
  3. 03CognitiveThe theory forms first, the evidence is sorted after
  4. 04LeverageCharges as bargaining chips, not descriptions
  5. 05RegulatoryA licensing matter, charged as a federal felony
  6. 06AttritionalThe process costs more than the sentence
  7. 07ProxyA private party aims the state at a rival

250pardons.com  ·  Every claim below is sourced

01

Form one

Political weaponization

A prosecution brought because of who the defendant is aligned with.

What it looks like

The Department brings charges that track a political affiliation, a public position, or proximity to someone in power. It is the form everyone means when they use the word, and the only one of the seven that requires proving what was in somebody’s mind.

Why we start here and move on

Several people on this site say plainly, on their own pages, that their prosecutions were brought in bad faith. We are not softening a word of that.

We are saying it is the one claim built to be unfalsifiable. A prosecutor who says he acted in good faith has given an answer that no procedure exists to test.

This is the only form on this page with no evidence section. That is not an oversight. It is the point.

Where this appears: Angelica Pacheco

02

Form two

Career weaponization

The conviction is worth something, and the acquittal is worth nothing.

What it looks like

Nobody is promoted for declining to charge. The measurable output of a prosecutor’s office is convictions, and no institution anywhere measures whether those convictions were right. A case that should have been dropped and a case that should have been won look identical on a résumé.

This is the form that needs no villain, which is exactly why it is the most dangerous one. Point an ordinary, well-meaning person at a scoreboard that counts only one thing, and they will get very good at that one thing.

I was not as interested in justice as I was in winning.

Marty Stroud III, the prosecutor who sent Glenn Ford to death row, speaking after Ford’s exoneration. Ford spent nearly thirty years in solitary confinement and died of lung cancer months after his release.

The evidence

Daniel S. Medwed, professor of law at Northeastern University, gave this its name in Prosecution Complex. His argument is that the pressures are institutional rather than personal — that conviction rates, office culture, and the ordinary machinery of advancement push even conscientious prosecutors toward securing and then defending convictions, including after evidence of innocence surfaces.

Angela J. Davis — professor at American University Washington College of Law, and for twelve years a public defender at the D.C. Public Defender Service — documented the structure that permits it. Her finding in Arbitrary Justice is that the perfectly legal exercise of prosecutorial discretion produces gross inequities, and that the courts, the Supreme Court included, have built a context that facilitates abuse and nonaccountability.

Neither of them argues that prosecutors are bad people. Both of them argue that it does not matter.

Daniel S. Medwed, Prosecution Complex: America’s Race to Convict and Its Impact on the Innocent (NYU Press, 2012)  ·  Angela J. Davis, Arbitrary Justice: The Power of the American Prosecutor (Oxford University Press, 2007). Not to be confused with Angela Y. Davis, the philosopher and activist.

Where this appears: Elizabeth Hernandez

03

Form three

Cognitive weaponization

The theory forms first. The evidence is sorted afterward.

What goes in

  • Fits the theory
  • Contradicts it
  • Fits the theory
  • Contradicts it
  • Contradicts it

What survives

  • Fits the theory
  • Fits the theory

Nothing was destroyed. It was reweighted.

What it looks like

An investigator settles on a suspect early. After that, everything consistent with the theory is credited and everything inconsistent with it is reinterpreted, discounted, or set aside as unreliable. You hand over the documents that prove the opposite of the government’s theory, and the documents change nothing. Not because anyone hid them. Because by the time they arrived, they were being read by people who already knew what they meant.

The evidence

Keith A. Findley of the University of Wisconsin Law School and Michael S. Scott published the definitive study of this in the Wisconsin Law Review in 2006. They define tunnel vision as the use of shortcuts that filter evidence selectively toward conviction, driven by confirmation bias, hindsight bias, and outcome bias. The prosecutor-specific version has its own term in the literature: conviction psychology.

Their central conclusion is not that some prosecutors are careless. It is structural, and it is worse.

They found the system in some ways embraces tunnel vision as a normative matter — that it demands or teaches it overtly, as a matter of policy or rule.

Keith A. Findley & Michael S. Scott, “The Multiple Dimensions of Tunnel Vision in Criminal Cases,” Wisconsin Law Review 2006, 291–397. Full paper

Where this appears: Brian Martinsen

04

Form four

Leverage weaponization

Charges used as bargaining instruments rather than descriptions of conduct.

100 federal defendants

Pled guilty    Went to trial

97.3% of federal defendants pled guilty in 2016. 97.2% in 2017. The share has not dropped below 97% in any year since.

What it looks like

The number of counts and the exposure they carry are set not by what a person did but by what it will take to make them sign. You are offered thirty months against a trial exposure of twenty years, and told to decide by Friday. Almost everyone signs. The system then records that choice as voluntary, which legally it is.

Federal sentence for the same crime

After a plea 1×
After trial 3× average

And sometimes eight or ten times higher. NACDL calls the differential coercive.

The evidence

Jed S. Rakoff — senior U.S. District Judge for the Southern District of New York, twenty-four years on the federal bench, and before that both a federal prosecutor and a defense lawyer — wrote in the New York Review of Books that the courtroom drama the Constitution imagines is a mirage: the system is almost entirely plea bargaining conducted behind closed doors, without judicial oversight, with the outcome determined by the prosecutor alone.

Nancy Gertner, herself a former federal judge and now a Harvard Law School professor, answered publicly that Rakoff was right and had not gone far enough. Her additions were threats, coercion, and prosecutorial power.

Two federal judges, writing in public, describing the system they administered.

NACDL, The Trial Penalty: The Sixth Amendment Right to Trial on the Verge of Extinction (July 2018) nacdl.org  ·  Jed S. Rakoff, “Why Innocent People Plead Guilty,” New York Review of Books, 20 November 2014.

Where this appears: Dr. Vichi Ganesh  ·  Ed Rosenberg

05

Form five

Regulatory weaponization

Conduct a licensing board or a regulator used to handle, charged as a federal felony.

The Department told its own prosecutors not to do this

7 Apr 2025The Deputy Attorney General issues “Ending Regulation by Prosecution,” directing prosecutors not to charge unlicensed money transmitting in digital-asset cases.
Four months laterRoman Storm is convicted of exactly that. As of publication he faces retrial.

What it looks like

A doctor’s billing judgment. A business owner’s paperwork. A software developer’s code. Conduct that a regulator would once have answered with a letter, a licensing action, or a civil suit, the Department charges as a federal felony — often under a statute written for something else entirely, and often without having to prove the defendant knew he was breaking any law at all.

The evidence

In May 2010 the Heritage Foundation and the National Association of Criminal Defense Lawyers jointly published Without Intent: How Congress Is Eroding the Criminal Intent Requirement in Federal Law. It was released at a Capitol Hill press conference sponsored by Representative Bobby Scott, a Democrat, and Representative Louie Gohmert, a Republican — then Chairman and Ranking Member of the House Crime Subcommittee. Edwin Meese, Attorney General under Ronald Reagan, spoke as well.

446Non-violent, non-drug criminal offenses proposed in the 109th Congress
57%Lacked an adequate guilty-mind requirement
4,450+Federal criminal statutes then on the books

A 2021 follow-up found 42 percent of the bills it analyzed still carried inadequate criminal intent requirements. Eleven years, one percentage point of movement for every three.

Reagan’s Attorney General and the criminal defense bar reached the same conclusion, on the same stage, sponsored by a Democrat and a Republican.

Walsh & Joslyn, Without Intent, Heritage Foundation and NACDL (May 2010) nacdl.org  ·  Smith & Pysno, Without Intent Revisited (2021) heritage.org  ·  DOJ, Office of the Deputy Attorney General, “Ending Regulation by Prosecution,” 7 April 2025.

Where this appears: Roman Storm  ·  Keonne Rodriguez  ·  Dr. Ron Elfenbein  ·  Dr. Sanjeev Kumar

06

Form six

Attritional weaponization

The cost arrives before any verdict, and no outcome returns it.

What it looks like

Legal fees that consume a retirement account. The government seizes assets before filing any charge. Years of waiting. A press release that outranks your own name in every search anyone will ever run on you, permanently. By the time a court reaches a judgment, the punishment has already been administered in full.

Civil forfeiture is the sharpest version

$82BTaken through forfeiture nationwide since 2000
D−Grade the Institute for Justice gives federal forfeiture law
100%Share of proceeds the seizing federal agency keeps

The Institute for Justice, a libertarian public interest law firm, has tracked this for two decades. Most laws let police and prosecutors take property and keep it permanently without ever proving the owner committed a crime, and then spend the proceeds themselves.

And most of it never reaches a courtroom. IJ found the majority of civil forfeiture cases end in default — no hearing, no arguments from both sides, no ruling on whether the seizure was lawful.

And the record is never corrected

The Justice Department’s own Justice Manual instructs that a press release issued before a finding of guilt should state that the charge is merely an accusation and that the defendant is presumed innocent. It contains no provision whatsoever for correcting or removing that release when the charges are dismissed, when the conviction is reversed, or when the defendant is acquitted.

Rules going in. Nothing coming out.

Malcolm M. Feeley, The Process Is the Punishment (Russell Sage Foundation, 1979) — the concept is his; the application here is ours  ·  Institute for Justice, Policing for Profit ij.org  ·  DOJ, Justice Manual 1-7.000 justice.gov/jm

Where this appears: Amy Nelson  ·  Chris & Erin Mazzei

07

Form seven

Proxy weaponization

A private party aims the state’s power at a rival, and the state carries out the strike.

What it looks like

The Justice Department does not discover most white-collar cases. It receives them. Complex fraud is expensive to investigate, so the Department depends on referrals — from regulators, from whistleblowers, and from companies that have already paid lawyers and forensic accountants to build the file. Whoever controls the intake shapes the enforcement.

Federal white-collar prosecutions, FY2024

4,332prosecutions of suspects
44involved charges against a corporation

In 99% of white-collar prosecutions since FY2004, the defendant was an individual rather than a business.

Referrals are selected, not automatic

Only 38 percent of white-collar criminal referrals result in prosecution, and just 6 percent of referred cases against businesses. A well-built package from a well-resourced institution moves. A complaint from one person does not.

And it is protected

Under the Noerr-Pennington doctrine, petitioning the government is shielded by the First Amendment even when the petitioner’s aim is to eliminate a competitor. The Supreme Court has held that joint efforts to influence public officials do not violate the antitrust laws even when intended to destroy competition. The narrow “sham” exception turns not on whether you meant to harm a rival, but on whether you actually wanted the government to act at all.

Handing a competitor to federal prosecutors, hoping they destroy him, is protected activity.

Transactional Records Access Clearinghouse, Syracuse University tracreports.org  ·  California Motor Transport Co. v. Trucking Unlimited, 404 U.S. 508 (1972)  ·  FTC staff report, Enforcement Perspectives on the Noerr-Pennington Doctrine.

Where a case on this site involves a private party whose complaint or referral preceded a federal prosecution, we publish the documented sequence of filings and dates. We do not characterize the motive of any private party.

Where this appears: Amy Nelson  ·  Ed Rosenberg  ·  Matthew Simpson

The proof problem

The same reasoning that convicts us is called speculation when we use it.

Here is the argument that sends people to prison. Read it twice.

Pointed at a defendant

Nobody has to hand you a confession. Look at the timing. Look at what he did and what he chose not to do. Look at what anyone in his position would have known. You may infer what he intended from the circumstances.

Convicts people every week

Pointed at the government

Nobody has to hand you a confession. Look at the timing — the indictment came three weeks after he filed the complaint. Look at what they did and what they chose not to do — eleven binders handed over, never opened. Look at what any prosecutor in that position would have known. You may infer what they intended from the circumstances.

Speculation

Federal juries hear a version of the first one constantly. Intent is an element of most federal crimes and there is almost never direct proof of what anyone was thinking — so the government does not offer proof. It offers a story assembled from the pattern of events, and the jury is instructed that this kind of reasoning carries the same weight as a document or a confession. It works because it persuades, not because it proves.

Three doors, all locked

You cannot sue. A prosecutor carries absolute immunity for prosecutorial decisions. Absolute, not qualified: the question of whether it was done in good faith is never reached, because the answer would not change the outcome.

You cannot be repaid. The one statute Congress wrote to reimburse a wrongly prosecuted defendant, the Hyde Amendment, requires showing the government’s position was vexatious, frivolous, or in bad faith. Bad faith. The state of mind nobody is permitted to prove.

And the profession does not step in. This is the door people assume is open. It is not.

California, 1997–2009 · Northern California Innocence Project

707 cases in which courts explicitly found prosecutorial misconduct
548 of those convictions upheld anyway — the misconduct ruled harmless
6 prosecutors publicly disciplined by the State Bar

Sixty-seven of them were found to have done it more than once. Three did it in four cases. Two did it in five.

It is not a California problem. The Center for Public Integrity reviewed more than 11,400 misconduct allegations nationally between 1970 and 2003 and found 2,012 in which the misconduct led to a dismissal, a reversal, or a reduced sentence. Forty-four prosecutors faced any disciplinary action at all. The Center for Prosecutor Integrity, pooling half a century of these studies, counts 3,625 documented cases of misconduct and public sanctions in 63 of them.

Apply to prosecutors the standard they apply to us — that intent may be read off conduct and consequence — and the courts already made the finding. Seven hundred and seven times. It cost six of them anything.

Kathleen M. Ridolfi & Maurice Possley, Preventable Error: A Report on Prosecutorial Misconduct in California 1997–2009, Northern California Innocence Project, Santa Clara University School of Law (October 2010)  ·  Center for Public Integrity, Harmful Error (2003)  ·  Center for Prosecutor Integrity, national registry prosecutorintegrity.org  ·  Imbler v. Pachtman, 424 U.S. 409 (1976)  ·  Hyde Amendment, Pub. L. No. 105-119, § 617, reprinted at 18 U.S.C. § 3006A note  ·  Pattern jury instructions on inferring intent from circumstantial evidence vary by circuit; consult the pattern instructions for the circuit of conviction.

Check your own case

Was any of this in your case?

Forty-two questions that sort what happened to you into these seven forms and tell you what you would need to document for each one.

It runs entirely in your browser. Nothing you enter is sent anywhere or stored, and we never see it. It cannot tell you whether you have a claim — it tells you where to look.

Start the checklist →

Why this framework exists

Some names appear under more than one form. That is not a filing error. The forms compound — a private referral produces a prosecution, the prosecution produces the pressure to plead, and the cost arrives long before any of it reaches a judge.

Nobody inside this system is required to answer for any of it. That is the whole finding. What the machine does with the inputs it is given is documented above. Who is left holding the result is documented in the cases.

Read the cases →