The Amerithrax Investigation: Anthrax, the FBI and a Contested Verdict

In the weeks after 11 September 2001, letters began arriving at newsrooms and Senate offices that had been packed with a fine, deadly powder. Inside envelopes postmarked Trenton, New Jersey, were spores of Bacillus anthracis, anthrax, milled fine enough to float into the lungs, together with notes in block capitals reading “DEATH TO AMERICA, DEATH TO ISRAEL, ALLAH IS GREAT.” Over the following weeks, five people died and at least seventeen fell ill. A postal worker, a hospital stockroom clerk, a photo editor in Florida and two others were killed by mail. The country, already reeling, now faced the possibility that the terrorist attacks had a biological sequel. The hunt that followed, codenamed Amerithrax, became one of the largest and most expensive investigations the FBI had ever run, and it ended nearly a decade later with a conclusion that a substantial part of the scientific community still does not accept. This is a case where the crime is beyond dispute and the solution is not, and holding those two facts together is the whole task.

What is not in doubt

Before the contested part, the confirmed part deserves to be stated plainly, because it is remarkable in itself. The anthrax attacks happened. Real letters carried real spores that killed real people, and the medical and forensic record of who died and how is solid. Robert Stevens, a photo editor at a Florida tabloid publisher, was the first to die, in early October 2001. Two postal workers at the Brentwood facility in Washington died after the contaminated mail passed through. The spores that hit the offices of Senators Tom Daschle and Patrick Leahy were prepared to a high standard, and the attack forced the closure and expensive decontamination of Senate buildings and postal facilities for months.

Two things about the material became clear early and shaped everything after. First, the anthrax was the Ames strain, a laboratory strain widely used in American biodefence research, which pointed the investigation inward, toward the small world of scientists with access to it, rather than outward toward a foreign cave. Second, the “Allah is Great” messaging was almost certainly misdirection. In the autumn of 2001 those notes fed a powerful current of suspicion that the letters were a second wave of Islamist terror, and even that they might be linked to Iraq, a theory that fed into the broader Iraq WMD intelligence failure then gathering force. Careful analysis of the spores and their preparation steered investigators away from a foreign state and toward a domestic scientist trying to look like a jihadist. That the perpetrator came from inside the American biodefence establishment is one of the few things about the case that nearly everyone came to accept.

The first suspect, and a wrecked reputation

The investigation’s early years produced a cautionary tale before they produced a suspect. In 2002 attention fastened on Steven Hatfill, a physician and bioweapons expert who had worked in the field and fit an early profile of who might have done it. In August 2002 Attorney General John Ashcroft publicly named Hatfill a “person of interest,” a phrase with no legal meaning that nonetheless functioned as a public accusation. Hatfill’s home was searched in front of cameras, he was followed openly by agents, one of whom ran over his foot, and his career was destroyed. He had done nothing. In 2008 the government settled his lawsuit for roughly $5.8 million, a tacit admission that it had ruined an innocent man in the glare of a national panic. That episode is worth remembering as context for what came next, because it demonstrated that the pressure to name a culprit could overwhelm the evidence, and it is one reason the eventual conclusion drew scepticism.

The second suspect, and a death before trial

The investigation’s forensic centrepiece was genetic. FBI scientists and outside laboratories developed novel techniques to trace the specific genetic signature of the attack anthrax, and they concluded that the spores derived from a particular flask of Ames-strain anthrax, labelled RMR-1029, held at the US Army Medical Research Institute of Infectious Diseases at Fort Detrick, Maryland. The custodian of that flask was Bruce Ivins, a microbiologist who had spent his career working on anthrax vaccines.

Reaching that flask took years of laboratory work that was, in its own right, a genuine scientific achievement. Investigators noticed that the attack anthrax was not genetically uniform: it contained a mixture of naturally occurring mutant variants, distinctive morphological types that showed up when the spores were cultured. The FBI, working with outside laboratories, developed assays to detect these specific variants and then screened a large repository of Ames-strain samples gathered from laboratories across the country and abroad, looking for the handful that shared the same combination of mutations. That painstaking process, more than any single clue, is what pointed at RMR-1029. The investigation ran for the better part of a decade, cost a reported figure well into the tens of millions of dollars, involved thousands of interviews and subpoenas, and pioneered the young field of microbial forensics more or less as it went. The scale of the effort is part of why its inconclusive ending was so jarring: an investigation that vast is expected to end in certainty.

As the investigation closed around him through 2007 and 2008, Ivins came under intense pressure. He had a documented history of mental-health difficulties, had made alarming statements to a therapist, and had behaved in ways the FBI read as consistent with guilt, including unusual late-night laboratory hours during the relevant period and attempts, investigators alleged, to mislead them about samples. In late July 2008, as prosecutors prepared to seek charges that could have carried the death penalty, Ivins died of an overdose of paracetamol, an apparent suicide. He never stood trial. In February 2010 the FBI and the Justice Department formally closed the Amerithrax case, concluding that Bruce Ivins had been the sole perpetrator, and released a lengthy investigative summary laying out the circumstantial and scientific case against a dead man who could no longer answer it.

Where the verdict comes apart

Here the case forks, and the fork is not the work of internet cranks. It runs straight through the scientific establishment. In 2008 the FBI itself asked the National Academy of Sciences, the most authoritative scientific body in the country, to review the microbial forensics that underpinned the conclusion. The Academy’s report, published in 2011, was measured and, for the FBI, uncomfortable. It found that the genetic evidence established an association between the attack spores and the RMR-1029 flask, but that the science could not definitively prove the attack material had come from that flask specifically, as opposed to other samples derived from it and distributed elsewhere. In the panel’s careful language, the scientific data were consistent with the FBI’s conclusion but did not independently establish it. The genetics, in other words, narrowed the field to a set of people with access to material of that lineage. They did not fingerprint Ivins.

Other reviews sharpened the doubt. A 2014 report by the Government Accountability Office criticised aspects of the FBI’s scientific approach and its failure to fully document how it reached its conclusions. Colleagues of Ivins publicly disputed his guilt, arguing that the RMR-1029 material had been shared with a number of other laboratories and that dozens of people, not one, had access to spores of that genetic lineage, and questioning whether Ivins had the equipment and skill to dry and mill the powder to the quality seen in the letters. The circumstantial behaviours the FBI cited, the late hours, the mental distress, were read by his defenders as the marks of a fragile, overworked man hounded toward suicide rather than of a mass murderer. None of this proves Ivins innocent. It means the case against him was never tested by the adversarial process that exists precisely to decide such questions, and that the one body of hard physical evidence, the genetics, was found by the nation’s premier science academy to be weaker than the FBI’s confident closure implied.

Why a contested verdict breeds theories, and why the honest answer is discomfort

An unresolved case of this magnitude is a natural home for conspiracy thinking, and Amerithrax has attracted its share: theories that Ivins was a scapegoat chosen to close an embarrassing file, that the real perpetrator remains protected, that the attacks were an inside job with darker sponsorship meant to grease the path to war. It is worth being clear about what is speculation. There is no documented evidence of a deliberate frame-up, and Ivins may well have been guilty; the FBI assembled a genuinely troubling circumstantial picture. But the structure of the case, an accused who died before trial, a headline forensic finding softened by the scientists asked to check it, an earlier innocent man publicly ruined, is close to a laboratory for manufacturing doubt. When the ordinary mechanism for reaching a verdict, a trial, never runs, the certainty that a trial produces is never available, and the space it would have filled fills instead with suspicion.

That pattern is not unique to Amerithrax. It is the same reason lingering distrust attaches to cases like the wrongful pursuit that preceded Ruby Ridge: when an agency has been seen to accuse the wrong person under pressure, its later confidence is discounted. The FBI’s treatment of Hatfill did lasting damage to its credibility on the very case it was trying to solve, because it proved the Bureau could be sure and wrong in the same breath.

What Amerithrax finally teaches is how much of our sense that a thing is “solved” depends on ritual rather than raw proof. A trial does more than determine guilt. It publicly demonstrates the reasoning, tests it against a defence, and lets a jury of citizens ratify a conclusion the whole society can then rest on. Ivins’s death removed that ritual, and the FBI’s unilateral closure could never substitute for it, no matter how thick the file. The result is a case that is, in the most literal sense, unfinished: a real and monstrous crime, a plausible but unproven culprit, and a forensic centrepiece that the country’s best scientists judged suggestive rather than conclusive. There is a further reason the case refuses to settle. The victims themselves, Robert Stevens, the postal workers Thomas Morris and Joseph Curseen, and the others, were ordinary people killed by mail addressed to strangers, and their deaths were never explained to them or their families by a verdict anyone could contest and lose. A trial would have owed them that accounting. Its absence leaves the harm without its answer, and a harm without an answer keeps generating questions long after the file is stamped closed. To insist the case is simply solved is to overstate what we know. To insist it is a cover-up is to invent what we do not. The honest place to stand is in the discomfort between, holding five deaths that certainly happened and a verdict that was never quite earned, and resisting the pull to resolve that tension with a stamp the evidence cannot support.