Hawaii man fights 34-year-old rape conviction based on debunked forensics

A bombshell 2009 report suggested widespread errors in the FBI's crime labs in the late 1980s and early 1990s, leading to convictions over potentially scientifically unsound evidence.

by Jeremy Yurow, Court House News, January 8, 2026

HONOLULU (CN) — Daniel Granillo walked into a Maui courtroom in 1990 facing kidnapping and sexual assault charges. He walked out with a 40-year prison sentence, convicted partially on the strength of FBI forensic testimony that a jury found compelling: hair and fiber evidence scientifically linking him to the victim.

Almost thirty-five years later, that evidence, and Granillo’s conviction, hangs in the balance.

On Thursday, the Hawaii Supreme Court considered if should Granillo receive a new trial or remain imprisoned based on forensic science the U.S. Department of Justice has since acknowledged “exceeded the bounds of science.”

FBI Special Agent Wayne Oakes told jurors in 1990 that hair found in Granillo’s car had been torn out at the roots and belonged to the victim, and that fibers from her clothing matched his car’s seat cover and carpeting.

The jury convicted Granillo of kidnapping, two counts of first-degree sexual assault, and attempted first-degree sexual assault stemming from a violent May 1989 attack on a woman in a Maui shopping center parking lot.

But in October 2017, federal authorities notified Maui prosecutors that Oakes’ hair comparison testimony was “inappropriate” and overstated what microscopic hair analysis could scientifically prove. The notification came amid revelations of widespread problems in FBI crime labs during the late 1980s and early 1990s, when analysts routinely exaggerated the reliability of hair comparison evidence.

Lower courts acknowledged the hair evidence was flawed but denied Granillo’s petition for a new trial, reasoning that Oakes’ fiber analysis remained valid and that other evidence, including the victim’s testimony and physical injuries, supported the conviction. The state Intermediate Court of Appeals upheld that denial, explicitly treating the fiber evidence as reliable.

The Hawaii Supreme Court accepted Granillo’s appeal specifically to examine that assumption.

“The FBI letter in this case only talks about the hair evidence,” Earl Partington, Granillo’s attorney, told the justices Thursday. “But as you can see from the 2009 report, the National Research Council concluded that fiber evidence is no better than hair evidence.”

Partington urged the court to take judicial notice of the landmark 2009 National Research Council report on forensic science, calling it the gold standard produced by some of the best scientists in the United States.

He also said that the report makes clear that fiber analysis suffers from the same scientific deficiencies as hair comparison. Both rely on subjective microscopic examination and can only determine whether samples belong to the same class of materials, not whether they came from a specific source.

The state, represented by Deputy Prosecuting Attorney Gerald Enriques, defended the conviction and resisted treating the fiber evidence as scientifically invalid.

“I think it’s premature, based on this current record, to say that the report would support a wholesale invalidation of the fiber evidence,” Enriques said.

He acknowledged the state has not identified studies contradicting the 2009 report’s conclusions about fiber analysis, but said prosecutors should have an opportunity to develop the record further.

The justices pressed both sides on whether the 2009 report’s findings about fiber evidence should apply to this case, and whether any distinction exists between the hair testimony the DOJ explicitly discredited and the fiber analysis Oakes also provided.

“This has really been discussed nationwide, right, in terms of the unreliability of hair and fiber,” Acting Chief Justice Sabrina McKenna said. “It talks about fiber, there’s 52 references to fiber because it’s a similar type of trace evidence like hair evidence, and it says the analysis is just like hair evidence. So why shouldn’t we take judicial notice of and note the unreliability of fiber evidence?”

The justices also questioned whether fiber evidence could still be considered valid when the expert witness and testing methods were identical to those used for the discredited hair analysis.

Enriques conceded he could not “stand here and say there wasn’t a possibility that it contributed to the conviction,” but maintained that even without the hair and fiber testimony, sufficient evidence existed to convict Granillo.

The justices also explored whether the forensic evidence was material enough to have influenced the jury’s verdict.

“The fiber evidence showed that the woman’s pants were undone,” Justice Todd Eddins said. “It showed that she was in that vehicle with her pants undone. That’s the difference with the fiber evidence and the hair evidence.”

Partington agreed the fiber testimony was part of the prosecution’s case and said it’s why its invalidation should warrant a new trial.

“There’s no way this was a harmless error,” he said. “The jury is going to look at forensic evidence and say this proves that the complainant was in the car, when we know from the government report it doesn’t prove any such thing.”

The case arrives at a moment of national reckoning with outdated forensic science. Pattern-matching disciplines like microscopic hair and fiber comparison, once presented to juries as near-certain proof, are now understood to lack the statistical foundations courts once attributed to them.

The FBI’s acknowledgment in 2015 that its analysts committed “widespread, systematic error” has prompted reviews of thousands of convictions.

The victim in the case, whose identity has not been publicly disclosed, reported that Granillo struck her in the parking lot, dragged her into his car, drove her to a beach, held her at knifepoint and repeatedly sexually assaulted and beat her before she escaped. The brutality of the attack resulted in Granillo receiving two consecutive 20-year sentences.

The Supreme Court panel — consisting of Acting Chief Justice Sabrina McKenna, Associate Justices Todd Eddins, Lisa Ginoza and Vladimir Devens, plus Circuit Judge Peter Kubota sitting by assignment due to a vacancy — gave no indication when it would rule.