Junk science drives wrongful convictions. That's not justice. | Opinion

Anyone who has watched an episode of "Law & Order" knows that prosecutors often rely on forensic experts dressed in white lab coats. They tell juries about microscopic hair comparisons, bitemark analysis, bloodstain patterns and other science-y stuff that's supposedly conclusive proof of guilt.

Except, too often, it's not.

Much of this so-called evidence is little more than the unsubstantiated opinion of someone with a stake in the outcome.

Yes, DNA is reliable (if collected properly). Too often, however, junk science is dressed up as factual evidence for jurors who don't know any better. In real life, courts accept phony forensics, often with troubling consequences.

Shaken baby syndrome and the unreliable science courts still lean on

In 2023, Nick Flannery was charged with assault and child endangerment after his infant son, Arlo, suddenly collapsed in a Cincinnati suburb. In the emergency room, Arlo's doctors found bleeding in the baby's brain and suggested it was caused by "shaken baby syndrome." Arlo had no bruises, no fractures, no skull injuries and no signs of neglect. His parents were the sole caregivers, with no history of abuse. There was no evidence whatsoever. It was the proverbial dog that didn't bark.

The allegation of shaken baby syndrome nevertheless set off a chain reaction.

The Flannery children were taken away. Independent medical experts reviewing the case found that a preexisting condition, which might have originated at birth and worsened over time, likely led to Arlo's bleeding and collapse.

Health records also showed unusual head growth and symptoms consistent with an ongoing medical condition, rather than abusive trauma. Intent on taking the case to a jury, prosecutors will argue that Arlo Flannery's injuries were from shaken baby syndrome, despite it being a long-disputed medical theory.

If convicted, Nick Flannery could face 12 years in prison.

The Flannery case is just one of many where unproven scientific theories are being used in criminal prosecutions. Robert Roberson was just days away from a scheduled Oct. 16 execution for the alleged murder of his 2-year-old daughter through what prosecutors were convinced was shaken baby syndrome.

On Oct. 9, the Texas Court of Criminal Appeals stopped Roberson's execution because state law required convictions to be reopened if previously used forensic evidence had since been discredited or proven scientifically inaccurate. Roberson will likely get a new trial, with the lead detective planning to testify about how unreliable shaken baby syndrome evidence is.

A week. That's how close the justice system came to an irreversible mistake.

Others, such as Josh Burns in Michigan and Andrew Roark in Texas, served years behind bars after being found guilty of shaking their kids to death. Their convictions were reversed by appellate courts due to the unreliable nature of the science used to convict them.

We have a path to rid our courts of junk science

To work, our justice system should depend on reliable, validated evidence rather than unproven theories and junk science.

Wrongful convictions rob innocent people of their freedom, drain millions in taxpayer dollars through protracted litigation and unnecessary incarceration, and force the government to pay for the substantial settlements required to redress when those convictions are overturned.

The Kentucky Innocence Project reports that junk science contributed to 52% of wrongful convictions in the cases they have reviewed, and that false forensic evidence was determined to be a contributing factor to 24% of all wrongful convictions nationally.

But the most significant impact of using phony forensics is probably on public confidence in our courts.

For our system of justice to work, state courts must follow the lead of federal courts by adopting the standards outlined in Federal Rule of Evidence 702, which require forensic testimony to be based on "reliable scientific principles and methods." This rule empowers judges to filter out flawed evidence before it taints the jury.

Unfortunately, vigorous validation happens infrequently in state courts. That's why it is more important than ever for states to adopt evidence-based standards to ensure that only credible and vetted information is given to juries.

Our Founding Fathers wanted to make sure courts were fair to every defendant. It is why we have jury trials in the first place. But the system only works when jurors are presented with trusted information rooted in sound science.

In "Law & Order," the right person is usually found guilty. In real life, that's not always the case. If we are to have true justice, judges must be empowered to set a high bar to screen out junk science and protect the innocent.

If states want to ensure they get it right, they should adopt the approach set out in Federal Rule of Evidence 702.

Van Jones is an attorney, criminal justice advocate and TV personality. David Safavian is an attorney and a conservative activist.

This article originally appeared on USA TODAY: Junk science drives wrongful convictions. That's not justice. | Opinion

Reporting by Van Jones and David Safavian, Opinion contributors / USA TODAY

USA TODAY Network via Reuters Connect

Copyright Reuters or USA Today Network via Reuters Connect

This story was originally published July 17, 2026 at 2:05 AM.