by Douglas Ankney

In March 2004, four commuter trains in Madrid, Spain, were bombed. Spain’s National Police recovered a blue plastic bag filled with detonators and traces of explosives. Forensic experts used the standard practice of fumigating the bag with vaporized superglue and staining it with fluorescent dye to reveal a blurry fingerprint. Spanish authorities submitted the print for a comparison search in the FBI’s Integrated Automated Fingerprint Identification System (“IAFIS”). IAFIS returned a possible match to Brandon Mayfield, a lawyer living in Oregon.

Three FBI experts then confirmed that the print from the blue bag matched Mayfield’s. A fourth expert, appointed by the court, also confirmed the match. The FBI arrested Mayfield. However, Mayfield had not been anywhere near Madrid during the bombings, and he did not even possess a current passport. Spanish authorities later confirmed that the print from the bag belonged to an Algerian named Ouhnane Daoud. Mayfield was released.

This international incident, involving a false-positive fingerprint identification by four of America’s top expert fingerprint analysts, shone a spotlight on fingerprint comparison analysis—commonly referred to as “friction-ridge fingerprint analysis”—at a time when the discipline was already under scrutiny. This article briefly examines (1) the standard method of fingerprint analysis and comparison, (2) the known flaws in fingerprint analysis and comparison, (3) the historical and current use of fingerprints as evidence in criminal trials, (4) courts’ reliance on the doctrine of finality to avoid reversing convictions tainted by unreliable fingerprint evidence, (5) known cases of false-positive identifications based on flawed fingerprint analysis, and (6) progressive steps forward in ensuring accurate fingerprint analysis and identifications.

FBI’s Statement on the
Brandon Mayfield Case

In a May 24, 2004, press release from the FBI, the agency apologized to Brandon Mayfield and his family for arresting Mayfield after wrongly matching his fingerprint to partial latent prints recovered from plastic bags containing detonator caps meant for use in the terrorist attacks on commuter trains in Madrid, Spain, in March 2004. The FBI explained that Spanish authorities had submitted images of the latent prints to the FBI for analysis. The images “were searched through the Integrated Automated Fingerprint Identification System (“IAFIS”).” An IAFIS search compares an unknown print to a database of millions of known prints. The result of an IAFIS search produces a short list of potential matches. A trained fingerprint examiner then takes the short list of possible matches and performs an examination to determine whether the unknown print matches a known print in the database.

“Using standard protocols and methodologies, FBI fingerprint examiners determined that the latent fingerprint was of value for identification purposes. This print was subsequently linked to Brandon Mayfield. The association was independently analyzed, and the results were confirmed by an outside experienced fingerprint expert.”

But shortly afterward, Spanish authorities made it known to the FBI that the recovered latent fingerprint did not belong to Mayfield. In a veiled attempt to cast blame on the Spanish authorities for the FBI’s blunder, the FBI stated that two fingerprint examiners were then sent to Madrid to compare “the image the FBI had been provided to the image the Spanish authorities had. Upon review, it was determined that the FBI identification was based on an image of substandard quality, which was particularly problematic because of the remarkable number of points of similarity between Mr. Mayfield’s prints and the print details in the images submitted to the FBI.”

But the FBI’s attempt to blame its blunder on Spanish authorities falls short since: (1) it was the FBI that found the print it initially received to “be of value for identification purposes,” and (2) it was the FBI that wrongly matched the print to Mayfield.

The Standard Method of Fingerprint Analysis and Comparison

The differences in fingerprints were observed at least as early as the first millennium before Christ. However, they were not proposed as a method for identifying individuals until 1880, when the idea was introduced by Dr. Henry Faulds. It was not until 1903 that the United States first implemented a system for fingerprinting.

Since 1911, when prosecutors first introduced them as evidence in criminal prosecutions, fingerprint analysis has been considered one of the most reliable methods in forensic science for identifying individuals and solving crimes. U.S. courts have routinely accepted fingerprints, and juries have considered them incontrovertible evidence.

“There are a lot of cases where the prosecution contends they have forensic ‘science.’ A lot of what they’re claiming to be forensic science isn’t science at all; it is mythology.” Linda Starr, Clinical Professor of Law at Santa Clara University in California and Cofounder of the Northern California Innocence Project

Fingerprints recovered from crime scenes or other areas of interest are classified into two categories: (1) patent prints, which are visible and (2) latent prints, which are invisible to the naked eye and require special techniques to be seen.

After a fingerprint is recovered from a crime scene, a computer algorithm compares the print to tens of millions of prints stored in a database. The algorithm produces a list of potential candidates ranked from most to least similar. A human examiner must then work through this list. The examiner, or “forensic technician,” compares the overall pattern and flow of the prints as well as the fine details in each, including ridge endings, bifurcations, contours, islands, dots, breaks, creases, pores, and enclosures.

The standard method of fingerprint identification employed by examiners in America is the four-stage “Analysis-Comparison-Evaluation-Verification” (“ACE-V”) method. An examiner must have a patent print or a latent print from the crime scene or area of interest and a known suspect print, commonly referred to as an exemplar, to conduct an ACE-V examination. The analysis phase encompasses a qualitative and quantitative evaluation of the fingerprint’s friction ridges at three levels of detail: (1) flow or direction of the ridges, (2) an examination of each individual ridge’s unique characteristics, and (3) a close examination of the pores of the ridge. While the first level of detail can be used to exclude but not to identify a print, it is a combination of the second and third levels of detail that may allow for either identification or exclusion. An examiner first analyzes the latent print and then the suspect print. If either of the prints is unsuitable for examination, the analysis ends.

If the prints are suitable for comparison, the examiner studies the friction ridge detail and determines whether the details match in similarity, sequence, and spatial relationship. There is no specific formula examiners use to determine whether there is a match. Indeed, no required number of similarities—often referred to as “points”—indicates a match, since it is both the quantity and quality of similarities that allow for identification. Seldom, if ever, do comparison prints match in every detail. Yet these dissimilarities are not dispositive either for or against a match.

The third phase in the ACE-V method, evaluation, is where the examiner forms a conclusion about the prints. The examiner may choose from one of three conclusions: (1) the prints are a match (known as “individualization” or “identification”), (2) they are not a match (known as “exclusion”), or (3) the result is inconclusive. It is important to observe that both the comparison and evaluation phases involve the exercise of subjective judgment by an examiner based on that examiner’s training and experience.

The last stage of the ACE-V method, verification, is when a second examiner analyzes the same latent and suspect prints in an effort to verify the first examiner’s conclusion.

The Known Flaws in Fingerprint Analysis and Comparison

Although fingerprints have been widely recognized as a means of identification for more than 100 years, there are still no definitive guidelines, and identifications are made using only a pictorial comparison. It is up to each examiner to decide whether they have identified a sufficient number of corresponding features to be confident that the two prints came from the same person. If so, this final “individualization” decision is logged into the computer system, where it is then typically “verified” by a second examiner who—depending on the jurisdiction—may or may not be blind to the initial examiner’s decision.

The Scientific Working Group on Friction Ridge Analysis, Study, and Technology (“SWGFAST”) is the recognized body charged with formulating guidelines for friction ridge impression examiners. However, even with SWGFAST’s three categories of identification, exclusion, or inconclusive, false identifications and exclusions are likely to occur because the expert analysis is highly subjective, especially since there are no official numerical thresholds to regulate the practice.

Moreover, even the most experienced forensic specialists face significant challenges, particularly when dealing with partial, smudged, or low-quality prints. FBI researchers revealed in a 2011 report that they had tested 169 experienced print examiners. The examiners were tested on 744 fingerprint pairs, of which 520 pairs contained true matches. Eighty-five percent of the examiners missed at least one of the true matches in a subset of approximately 100 pairs each examined. The examiners also demonstrated inconsistency. In a subsequent study seven months later, the researchers brought back 72 of those examiners and gave them 25 of the same fingerprint pairs they had seen earlier. The examiners changed their conclusions on about 10% of the pairings.

Some common causes of errors in forensic fingerprint identifications are:

Partial Prints: Most prints gathered at crime scenes aren’t complete. Partial samples are commonly used, and there can be a great deal of variability in ridge details.

Smudges and Distortion: Prints can be blurred or distorted by the hand movement of the owner, the character of the surface, moisture, or other factors. These elements can further complicate fingerprint collection.

Human Error: Judgment plays a part in fingerprint analysis, and studies have shown that even well-trained and experienced analysts can make mistakes when evaluating prints for matches.

There are special challenges associated with latent fingerprint analysis (prints that are invisible to the naked eye and require special techniques to be revealed). These include surface types, moisture, temperature, other environmental conditions, and the tendency of prints to degrade with time and subsequent contact.

In 2016, the President’s Council of Advisors on Science and Technology (“PCAST”) found that, of all the studies of latent fingerprint analysis that had been conducted, only two were properly designed. Those two studies revealed disturbingly high rates of false-positive identifications: 1 in 18 and 1 in 30.

Even the handful of studies of fingerprints that were arguably biased to favor fingerprint examiners showed a troubling pattern of errors. Since 1995, Collaborative Testing Services, a company that evaluates the reliability and performance of fingerprint labs, has administered an annual, voluntary test. It sends fingerprint labs a test that includes 8 to 12 pairs of prints that examiners confirm or reject as matches. The bias in these tests is that the pairs usually consist of complete, not partial, prints, making identifications easier than the real situations examiners face. Nevertheless, the error rate has varied from 3% to a shockingly high 20%.

While research has now convincingly shown that fingerprint ridge patterns vary greatly among individuals (even among identical twins) and there is almost no variation in these patterns over time, there is still no scientific basis for concluding that a print must have been left by a specific person. The most that may be concluded from fingerprint analysis is that two prints “have many corresponding features with no differences that would indicate they were made by different fingers.”

The fundamental problem is the lack of underlying statistical evidence. Genetic evidence provides a good model for comparison. Developed in 1984 and first introduced in a U.S. court in 1987, scientists and lawyers subjected the technique to years of scientific scrutiny, along with almost a decade of court challenges, before it became accepted evidence. With DNA analysis, examiners identify and compare short segments of DNA to make a match. The number of segments required to verify a match is generally agreed to be 13. In addition to having established procedures for analyzing evidence, experts have calculated the odds that two people could share the same DNA in all 13 segments. These odds vary slightly based on the prevalence of certain DNA patterns among different ethnic groups but are in the tens of millions to one against two people sharing all 13 segments.

Unlike DNA evidence, fingerprinting was adopted before the Supreme Court decided that attorneys and expert witnesses must prove that evidence is scientific and reliable. With regard to fingerprint comparison and analysis, “[t]here has been a hundred years of precedence, not a hundred years of data,” said consultant Ralph Haber, who studies the reliability of forensic evidence.

While everyone on the planet may have a unique set of prints, the real question is whether experts can accurately connect the prints collected by crime scene investigators to the right person. Crime scene prints typically consist of only about 20% of the fingertip, and they are often smudged. Examiners link a partial print from a crime scene to a whole one taken from a suspect by matching particular characteristics of the fingerprint. “They compare the overall print pattern and other ridge characteristics, including the width of the ridges and the spacing of oil pores,” according to Ed German, a fingerprint examiner with the U.S. Army. “But examiners primarily rely on matching points on both prints where ridges end, bifurcate, or change direction.”

Examiners generally conclude a crime scene print came from a suspect after matching from 3 to 16 points (FBI examiners found 15 points on Mayfield’s print). But there are no standards or requirements on the number of points that must be matched. Disturbingly, each lab—and sometimes each examiner—determines the number needed. It is known that each practitioner tends to create and follow their own thresholds for identification, even though the suggested average threshold for identification is a minimum of 8 to 9 ridge feature matches.

Fingerprint examiners are fond of touting the permanence and uniqueness of fingerprints. However, they are unaware of the odds that two people could share a given number of fingerprint characteristics. With no clear guidance on how much relevant weight to give the various print characteristics—like point matches, ridge width, and spacing of oil pores—German argues that it is impossible to attach probabilities to print identifications.

Many experts are persuaded that probabilities are unnecessary since examiners would not make or confirm an identification unless they were certain of it. But when three of the most experienced FBI examiners confirm a mistake, as they did with Mayfield’s prints, the argument collapses. Other print proponents argue that, despite occasional human errors, the method is infallible. But critics like Simon Cole, a legal historian who has testified in many court challenges, rightly point out that this is a useless distinction. Regardless of the reason, fingerprint identifications are sometimes wrong.

Additionally, due to the extensive reliance on human judgment in fingerprint analysis, an examiner’s conclusions may be influenced by “cognitive bias.” Cognitive bias occurs when the results of an examination are skewed by nonscientific factors. For example, in a recent study, five fingerprint expert examiners were told they were comparing Brandon Mayfield’s fingerprint with the fingerprint recovered in the Madrid train bombings. These experts were asked if they would have made a false-positive match like the FBI had done. In actuality, each expert was given fingerprints from a different case they had examined earlier in their routine casework and had concluded matched. Four of the five experts (80%) changed their opinion due to their expectation that the prints did not match.

Another study involved the use of an automated fingerprint identification system (“AFIS”). The AFIS provided ordered lists of prints that were likely matches, beginning sequentially with the print most likely to match at the top and moving down to the print least likely to match at the bottom. The study found that expert examiners are more likely to wrongly identify a print near the top of the list as a match and fail to correctly make an identification if the print is lower on the list.

In yet another example of cognitive bias, a panel of outside print examiners convened by the FBI to review the Mayfield case found that a supervisor made the initial identification, and lower-ranking examiners, when asked to confirm or reject their boss’s work, felt pressured to confirm. FBI supervisors making the initial identifications is not unusual, according to Alan McRoberts, a retired Los Angeles County Sheriff’s Department examiner and chair of the review panel. Other agencies do it as well. But while this practice resulted in one wrongful arrest, the panel did not recommend the FBI review other cases. “As a committee, I don’t think we discussed that in particular,” said McRoberts.

Another troubling test was conducted by the FBI. When defense lawyers questioned the reliability of fingerprint identifications during Byron Mitchell’s armed robbery trial in 1999, the FBI responded by sending two prints taken from a getaway car and Mitchell’s prints to 53 crime labs to confirm the agency’s identification. Of the 39 labs that returned their results, nine (23%) concluded that Mitchell’s prints did not match those from the car. Nevertheless, the judge rejected the defense’s challenge and accepted the fingerprint evidence. Mitchell was found guilty and remains in prison. The FBI has not repeated the experiment.

There are several reasons to question the ability of fingerprint examiners to identify the donor of a crime scene print “to the exclusion of all others” via friction ridge analysis. First, invalidated and improper forensic evidence has contributed to almost half of the wrongful convictions (exonerated by DNA) identified by the Innocence Project as of 2016.

In 2009, the National Academy of Sciences (“NAS”) produced a landmark Justice Department-funded report, Strengthening Forensic Science in the United States: A Path Forward (“NAS Report”). It concluded that, “[w]ith the exception of nuclear DNA analysis … no forensic method has been rigorously shown to have the capacity to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual or source.”

The NAS Report acknowledged that friction ridge analysis has “long served as a valuable tool, both to identify the guilty and to exclude the innocent.” It also gave some support to the discipline’s ability to engage in individualization: “Because of the amount of detail available in friction ridges,” the NAS speculated, “it seems plausible that a careful comparison of two impressions can accurately discern whether or not they had a common source.” The NAS Report additionally agreed that “[s]ome scientific evidence supports the presumption that friction ridge patterns are unique to each person and persist unchanged throughout a lifetime.”

However, the NAS Report criticized friction ridge analysis in several areas. It found that ACE-V “is not specific enough to qualify as a validated method” because it “does not guard against bias; is too broad to ensure repeatability and transparency; and does not guarantee that two analysts following it will obtain the same results.” Thus, an examiner is not “proceeding in a scientific manner or producing reliable results” when applying ACE-V. The NAS Report also recommended that examiners be more thorough in documenting their analysis and unequivocally stated that claims of a “zero error rate” are “unrealistic.”

“It’s just a nightmare knowing that someone who’s innocent can be picked up off the street and held. That scares me now. It’s like I’m walking on eggshells. I try to cover my tracks for everywhere I go.” Dwight Gomas, incarcerated for 17 months for a robbery he did not commit after he was wrongly identified as the source of fingerprints left at the crime scene

The Historical and Current Use of Fingerprints as Evidence in Criminal Trials

The term “forensic science” implicitly refers to the use of a scientifically based discipline as it intersects with and provides evidence for legal proceedings. The Federal Rules of Evidence (“FRE”) set out the framework within which evidence is admitted into court. The primary rules that apply to expert witnesses are FRE 702 (Testimony by Experts) and FRE 703 (Bases of Opinion Testimony by Experts):

A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if:

(a) the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or determine a fact in issue;

(b) the testimony is based on sufficient facts or data;

(c) the testimony is the product of reliable principles and methods; and

(d) the expert has reliably applied the principles and methods to the facts of the case. Rule 702, Testimony by Expert Witnesses, Rules of Evidence for United States Courts and Magistrate Judges (2020) (FRE 2020)

An expert may base an opinion on facts or data in the case that the expert has been made aware of or personally observed. If experts in the particular field would reasonably rely on those kinds of facts or data in forming an opinion on the subject, they need not be admissible for the opinion to be admitted. But if the facts or data would otherwise be inadmissible, the proponent of the opinion may disclose them to the jury only if their probative value in helping the jury evaluate the opinion substantially outweighs their prejudicial effect. Rule 703, Bases of an Expert’s Opinion Testimony, FRE 2020

The SWGFAST postulates that forensic friction ridge impression examination is “an applied science based upon the foundation of biological uniqueness, permanence, and empirical validation through observation.” This contention is based on a belief in (1) the uniqueness of friction ridge patterns and (2) the persistence of the friction ridge arrangements.

It is known that fingerprints or fingermarks refer to the impressions left on a surface due to contact with an individual’s fingers. The unique ridge patterns present on the finger pads of humans—made up of various combinations of loops, whorls, and arches, known as “friction ridges”—create these impressions or prints. The friction ridges form on the bottom of an individual’s hands and develop between the 9th and 24th weeks of embryonic development.

Because the ridges develop so early in life, the movement and pressure inside the womb vary the patterns. Since it is unlikely that both of these conditions are identical for different embryos, each fingerprint pattern is unique to a particular person, with the overall fingerprint patterns differing from finger to finger and person to person, including siblings and twins. Based on this conclusion, fingerprints are used as a means of identification.

Proponents of forensic fingerprint analysis and identification argue that, without an understanding of the biological aspects underlying the formation of friction ridges prenatally, experts would never be justified in concluding that one individual was the source of an impression to the exclusion of all other possible persons. The SWGFAST position thus supports the claim that forensic friction ridge impression examination is scientific.

However, it is equally true that friction ridge impression examinations require specialized technical knowledge to reach reliable conclusions. Consequently, forensic friction ridge impression examinations can be proffered as meeting any or all three of the prongs contained in FRE 702(a).

As to the “sufficiency” requirement of FRE 702(b), proponents of the discipline claim, “It has been established by sound and repeated studies that friction ridge examination evidence permits the uncontroverted association of an individual with a particular scene or object. If the scene or object is part of a crime, the individualization evidence would certainly offer a logical connection to a case, permitting a jury to draw conclusions about the guilt or innocence of the individualized person.”

The requirements of FRE 702(c) are addressed by SWGFAST’s established and approved methodology—the ACE-V method—designed to lead to reliable and verifiable conclusions if the prescribed methodology is followed by a competent examiner.

Finally, FRE 702(d) is based on the expert witness’s testimony as to whether they applied the principles of FRE 702(c) to the instant case in arriving at their conclusion.

However, deviation from the standard SWGFAST methodology and applied principles may be warranted if the nature of a particular case falls outside the normal course of investigation. Such a deviation “requires significant thought and experience on the part of an examiner, but the justification for the deviation must always be clearly documented in the examiner’s notes.”

If, during the course of the investigation, the fingerprint examiner uncovers or discovers material evidence that is exculpatory to the accused, the evidence must be disclosed to the prosecution, the court, and the defense. Brady v. Maryland, 373 U.S. 83 (1963). Furthermore, the U.S. Supreme Court ruled in Giglio v. United States, 405 U.S. 150 (1972) that “the government is constitutionally required to disclose any evidence favorable to the defense that may impact the defendant’s guilt or punishment, including any information that may bear on the credibility of its witnesses, even if the defendant fails to request such information.”

According to United States v. Henthorn, 931 F.2d 29 (9th Cir. 1991), this includes the government’s “duty to review the personnel files of its testifying officers and to disclose to the defense any information that may be favorable to the defendant and meets appropriate standards of materiality. Obviously, this is information that would go to the qualifications of the experts. Such matters as past errors, required training, or any actions that may reflect on integrity or credibility are subject to this ruling.” While Henthorn is not a Supreme Court opinion, it is widely followed in other jurisdictions.

The development of the federal judiciary’s rules for the admission of scientific evidence has a rather lackluster beginning. In the first quarter of the twentieth century, William Marston developed a polygraph machine and was prepared to testify, based on the results of an examination using the polygraph machine, that murder suspect James Adolphus Frye spoke truthfully when he denied any knowledge of the crime. But the trial judge refused to let Marston testify. On appeal, the Court stated:

“Just when a scientific principle or discovery crosses the line between experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone, the evidential force of the principle must be recognized, and while the courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.” Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).

The Frye Court concluded that the polygraph test had not gained general acceptance in the fields of human physiology and psychology. The decision was largely ignored by the majority of courts until the 1960s when, as a result of massive federal assistance, crime labs began to flood courtrooms with myriad types of expert testimony on newly developed “scientific methods,” such as bite mark comparisons, questioning with truth serum, voiceprints, and so on. Known as the “Frye Test,” whether an expert’s method “had gained general acceptance in the particular field in which it belong[ed]” became the standard for the admittance of the expert’s testimony in many courts.

However, in federal courts, the Frye Test was replaced in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). The Daubert Court explained that trial courts are “gatekeepers” tasked with ensuring that only “relevant” and “reliable” evidence is admitted. The Court determined that Frye’s “rigid” general acceptance standard is “incompatible with the Federal Rules of Evidence.” Daubert charged judges to examine the principles and methodology of proffered scientific evidence, not just whether its conclusions were generally accepted in the scientific community. The Daubert Court articulated five non-exhaustive factors for trial courts to consider when deciding whether to admit expert opinion testimony based on scientific knowledge: (1) whether a theory or technique can be (and has been) tested, (2) whether it has been subjected to peer review, (3) whether there are standards controlling the technique’s operation, (4) the known or potential error rate, and (5) whether it is generally accepted within the relevant scientific community. And in Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), the Supreme Court clarified that the Daubert factors apply to all expert testimony as defined by FRE 702.

The first serious Daubert challenge to the admission of expert testimony with regard to fingerprints occurred in United States v. Mitchell, 365 F.3d 215 (3d Cir. 2004). In affirming the district court’s admission of the evidence, the Third Circuit explored each of the Daubert factors.

The Mitchell Court explained, “[t]estability refers to ‘whether the premises on which fingerprint identification relies are testable—or, better yet, actually tested.’” The Court concluded that the premises that friction ridge arrangements are unique and permanent, and that a positive identification can be made from fingerprints containing sufficient quantity and quality of detail, were testable and had been tested in several ways. In that regard, the Court referred to the FBI’s AFIS computer comparison of 50,000 left-sloped patterns against a database of another 50,000 sets of tenprint, a process involving 2.5 billion comparisons. The experiment showed there were no matches of prints coming from different digits.

The Mitchell Court was not impressed by the government’s argument that the verification step of “ACE-V” constituted peer review. The Court was of the view that the examiners had developed an “occupational norm of unanimity” that discouraged dissent. Nevertheless, the Court concluded that “the ACE-V verification step may not be peer reviewed in its best form, but on balance, the peer review factor does favor admission.”

As to error rates, the Court shifted the burden to the defense to prove false-positive identifications. While the error rate had not been precisely quantified, “the Court was persuaded that the methods of estimating it showed it to be very low.”

Regarding maintenance of standards, the Court found this to be “lacking in some measure.” The procedures of ACE-V were deemed “insubstantial in comparison to the elaborate and exhaustively refined standards found in many scientific and technical disciplines” and found that this factor did “not favor admitting the evidence.”

The Court concluded that “most factors support (or at least do not disfavor) admitting the government’s” evidence. Additionally, the Court noted: “Experts with diametrically opposed opinions may nonetheless both have good grounds for their views, and a district court may not make winners and losers through its choice of which side’s experts to admit when all experts are qualified.”

In United States v. Llera Plaza, 179 F.Supp.2d 492 (E.D. Pa. 2002) (“Llera Plaza I”), Judge Louis H. Pollak ruled that the ACE-V method generally used to arrive at match or non-match conclusions does not meet the first three Daubert factors and only met the general acceptance factor in the technical, as opposed to the scientific, community of fingerprint examiners. Based on that finding, he allowed fingerprint experts for both prosecution and defense to testify to all of the examinations they had performed in an individual case, yet he would preclude them from testifying that the latent and inked prints were, or were not, from the same person.

But after a hearing on the Government’s motion for reconsideration, Pollak said, “I have changed my mind.” By applying the legal mandates announced in the Daubert and Kumho Tire cases, Pollak concluded that (1) judicial notice would be taken of the permanence and individuality of friction skin (fingerprint) patterns and (2) experts in the field would be permitted to express their opinions on a match of two impressions. He arrived at this conclusion after he heard or read the explanations of law enforcement-trained examiners and university-based scientists in genetics, histology, and fetal development regarding the biological and physiological factors that result in ultimate pattern uniqueness during the prenatal development of friction skin.

Courts have ruled that fingerprint evidence alone can be sufficient to sustain a conviction. In Stacy v. State, 292 P. 885 (Okla. Crim. App. 1930), the defendant’s fingerprints were found on a door of a vault that was breached. The defendant argued that the fingerprint evidence alone, without corroboration by other facts or circumstances, was insufficient to sustain his conviction. The Stacy Court explained:

From an examination of the authorities cited and others, it appears that an allusion to fingerprint impressions for the purposes of identification is referred to in writings as early as 600 A.D., and they are traced back to a period some 100 years before Christ…. We have no doubt but that the finding of the fingerprints of the defendant on the door of the vault, with the further proof that the defendant did not have access to and had not been at the place burglarized so that the prints could be accounted for on any hypothesis of his innocence, is a circumstance irresistibly pointing to his guilt.…

In 2009, defendants began challenging the admission of fingerprint evidence based on the deficiencies in the discipline of friction-ridge analysis outlined in the NAS Report. In United States v. Rose, 672 F.Supp.2d 723 (D. Md. 2009), the Court ruled that precedent, the general acceptance of the ACE-V method in the scientific community, and the lack of evidence contradicting the Government’s assertion that misidentifications were extremely rare all favored admission. While the Court acknowledged that the NAS Report made clear that there was no “available scientific evidence of the validity of the ACE-V method,” the NAS Report “did not conclude that fingerprint evidence was unreliable such as to render it inadmissible.” The Court emphasized Daubert’s view that “[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof” are the “appropriate methods of attacking perceived flaws in admissible scientific or technical evidence.”

In Commonwealth v. Gambora, 933 N.E.2d 50 (Mass. 2010), the defendant used the NAS Report to challenge evidence that “matched” his fingerprints to latent prints found at the crime scene. The Massachusetts Supreme Judicial Court considered the concerns raised in the NAS Report, including “the subjective and impure nature of the ACE-V, the lack of a quantifiable error rate, and the need for more extensive research to underpin the discipline.”

Regarding cross-examination as a means to expose the flaws in forensic fingerprint analysis, the Court recognized some limitations on its impact:

While we normally leave the humbling of inflated opinions to cross-examination, there is a danger that the mystique of fingerprint identification, which has had a captivating hold on the criminal justice system and society at large for more than one hundred years, is such that cross-examination may not be enough to rectify the effect of a fingerprint expert’s use of such terms as ‘individualized,’ ‘absolute,’ and ‘match’ when testifying, as opposed to presenting the testimony as his or her ‘opinion’ that the latent fingerprints are the defendant’s.

Nevertheless, the Court admitted the evidence. It recognized the “confusing conclusions” of the NAS Report: “As our discussion of the NAS Report reflects, there is tension in the report between its assessments that, on the one hand, ‘it seems plausible that a careful comparison of two impressions can accurately discern whether or not they had a common source,’ but that, on the other, ‘merely following the steps of ACE-V does not imply one is proceeding in a scientific manner or producing reliable results.’”

Apparently, in recognizing that cross-examination of fingerprint experts is insufficient to impress upon jurors the deficiencies of friction-ridge analysis, the Gambora Court sought to limit such testimony: “[B]ased on the NAS Report, we can say this much at the present time: Testimony to the effect that a latent print matches, or is ‘individualized’ to, a known print, if it is to be offered, should be presented as an opinion, not a fact, and opinions expressing absolute certainty about, or the infallibility of, an ‘individualization’ of a print should be avoided.”

Similarly, in United States v. Alan, 748 F.Supp.2d 531 (E.D. Va. 2010), while addressing the defendant’s motion to exclude fingerprint evidence that allegedly linked him to an arson fire, the Court stated that “[t]he absence of a known error rate, the lack of population studies, and the involvement of examiner judgment all raise important questions about the rigorousness of friction ridge analysis.”

Regarding the lack of a known error rate, the Court noted: “[W]hile fingerprint experts sometimes use terms like ‘absolute’ and ‘positive’ to describe the confidence of their matches, the [NAS Report] has recognized that a zero-percent error rate is ‘not scientifically plausible.’” However, once again, the Court ruled that the defendant’s challenge “was appropriate for cross-examination, and not grounds for exclusion.”

In State v. Sheehan, 273 P.3d 417 (Utah Ct. App. 2012), the defendant argued: (1) “the trial court abused its discretion when it failed to hold [an admissibility] hearing to determine whether to admit the State’s expert testimony that a palm print found at the scene matched Sheehan’s palm print” and (2) the trial court erred in its exclusion of Sheehan’s expert’s testimony as well as its limitation of “Sheehan’s cross-examination of the State’s experts.” The appellate court upheld the trial court’s decision to admit the State’s evidence but found that the trial court erred in restricting Sheehan’s cross-examination of the State’s experts: “In depriving a person of life or liberty … due process [is] … [a] fair opportunity to submit evidence, examine, and cross-examine witnesses.”

“Thus, the trial court’s legal determination that the State’s expert testimony was admissible did not allow the court to then impinge on the jury’s role as fact finder by excluding the evidence that Sheehan may have used to challenge the credibility and weight of the State’s expert testimony.” Without appropriate justification, “exclusion of this kind of exculpatory evidence … deprives a defendant of the basic right to have the prosecutor’s case encounter and survive the crucible of meaningful adversarial testing.”

In United States v. Stone, 848 F.Supp.2d 714 (E.D. Mich. 2012), the defendant moved to exclude expert testimony based on the NAS Report. The Court rejected Stone’s motion, finding that “concerns about the risks of error, such as false-positive identifications, go to the weight of the evidence and can be explored on cross-examination and/or through presentation of competing evidence.”

“We continue to have cases popping up, sort of right and left, in which they are doing it wrong.” Psychologist Gary Wells from Iowa State University, Ames

After Daubert resulted in a revision of FRE 702 that requires expert evidence to be based on “reliable principles and methods” that are “reliably applied” to the facts of a case, it was believed that trial courts would be more rigorous in their role as gatekeepers. And in light of the advent of modern DNA testing that inadvertently revealed the shortcomings of other disciplines of “forensic science,” including fingerprint analysis, it was believed those disciplines would be held to Daubert’s standards.

But the opposite is true. “As Simon Cole has observed, judges have grandfathered in latent fingerprint evidence based on its longstanding use, and not based on any evidence that it is in fact reliable. State judges have frequently done the same. For example, a recent ruling by an Arizona appellate court emphasized: “[O]ur supreme court has sustained convictions based solely on expert testimony about fingerprint or palm print evidence because the evidence is sufficiently reliable.” State v. Favela, 323 P.3d 716, 718 (Ariz. Ct. App. 2014).

Courts’ Reliance on the Doctrine of Finality to Avoid Reversing Convictions Tainted by Unreliable Fingerprint Evidence

While these decisions reveal that courts admit “shaky” fingerprint evidence on the assumption that flaws in the evidence will be uncovered during cross-examination, the facts demonstrate this assumption is a fallacy. Moreover, guilty verdicts after trials where shaky fingerprint evidence has been admitted are seldom overturned on appeal.

Postconviction relief grounded on unreliable fingerprint evidence is stymied by the principle of “finality.” This principle developed out of a “taxonomy” described by Professor Paul Bator in his seminal 1963 article, “Finality in Criminal Law and Federal Habeas Corpus for State Prisoners.” Bator laid the intellectual groundwork for the Supreme Court’s post-trial review jurisprudence and has been cited in hundreds of law review articles and court opinions. He argued that the finality of criminal judgments serves important interests that are harmed by the expansion of post-trial rights. Bator argued that, because we can never be 100% certain that no error of law or fact was made during trial (or appellate) proceedings, we must impose an end to litigation at some point, or else the case could conceivably go on indefinitely.

Professor Carrie Sperling describes “finality” as follows:

[P]ost-conviction procedures grow out of a strong tradition that values the finality of criminal convictions. Finality in the criminal law context means that the case is over, with no avenues remaining to challenge the conviction. Finality assigns guilt, puts the case to rest, and assures all parties that it will not be reopened. Assuming that the criminal process is error-free, finality serves the ends of justice. Theoretically, once the system convicts the right person and assesses the right punishment, society is better off when prosecutors, law enforcement professionals, defense attorneys, and judges move on to other concerns.

“Finality” has come to mean “shorthand for a collection of interests scholars assume are furthered by any restrictions on review. Finality assumes that providing defendants broader postconviction rights harms these society-desired interests. Consequently, when considering appeals, judges must balance society’s interests in finality against the rights of defendants.”

Unsurprisingly, courts have fully and eagerly embraced the principle of finality (especially for criminal cases). Judges and scholars routinely assert that limiting defendants’ postconviction avenues benefits society. However, trial courts’ use of finality to rationalize decisions to allow the admission of fingerprint identification evidence arguably overlooks the limitations of the adversarial system.

One glaring limitation lies in the inability of lawyers and jurors to grasp the significance of errors in forensic scientific methods. Relying on a lawyer’s ability to expose errors during cross-examination of the expert is problematic for many reasons. First, lawyers are generally not scientists. As Professor David L. Faigman stated, “Not only do they not have training in the particular [scientific] subject, they have a more profound disability: most … lack the ability to judge whether proffered research is good science, bad science, or science at all.”

Similarly, Professor Fredric Lederer noted, “[L]awyers generally lack significant scientific training. This educational deficiency often places lawyers at a disadvantage when confronted with scientific evidence…. [L]awyers … often fail to ask the right questions and uncritically accept scientific assertions.”

The NAS Report echoed this shortcoming as well, stating more than 10 times that “lawyers and judges often have insufficient training and background in scientific methodology, and they often fail to fully comprehend the approaches employed by different forensic science disciplines and the reliability of forensic science evidence that is offered in trial.”

An example demonstrating the dysfunction within the adversarial process regarding forensic science is a report from the NAS evaluating voice spectrography used to identify suspects. While the report was highly detailed and immediately adverted to by the FBI, “[l]awyers in trials around the country failed to find and bring the report to the attention of judges, judges failed to find the report, and several courts that clearly knew of the report failed to learn from it.” Professor Michael Saks summed it up: “[T]he adversary process failed to motivate lawyers to find and offer the most important evidence on the subject at issue.”

Second, defense counsel often has limited resources. While relying on defense counsel to effectively cross-examine experts to expose faulty forensic science, “reducing the number of trial errors would generally require attorneys to spend more time and resources representing each client.” Forensic experts are generally expensive to hire, and their applications can be time-consuming. Trial counsel cannot magically create these resources (along with adequate scientific knowledge) simply because they know postconviction review is limited.

And most jurors, like lawyers, are not scientists. As a consequence, many jurors are unable to competently engage with scientific evidence or determine the appropriate weight to afford specific evidence. There is sufficient evidence supporting the conclusion that jurors consider forensic evidence especially critical to their ultimate decision about guilt; that they have a thirst for scientific evidence; and that they expect to see it, especially in cases where the majority of evidence is circumstantial.

Additionally, further research suggests jurors give great deference to statements made by experts, and the impact of an expert’s statement is unlikely to be undone either through cross-examination or rebuttal witnesses. Surprisingly, it is also suggested that jurors feel more inclined to credit an expert’s testimony when subjected to vigorous cross-examination and more skeptical about the reliability of a defendant’s rebuttal evidence.

Indeed, Individualization Claims in Forensic Science: Still Unwarranted, Jonathan J. Koehler & Michael J. Saks, 75 Brook. L. Rev. 1187 (2010), concluded:

One might have expected an explication of the [forensic science] examination process, emphasizing the guesswork involved, would have a sobering effect on fact finders, but it appears instead to lead fact finders to be more impressed by the examination. Similarly, since most jurors begin with an exaggerated view of the nature and capabilities of forensic identification, one might expect that information explicitly informing fact finders about the limitations of the expertise would temper the jurors’ inferences. Such information had little effect on jurors’ judgments.

Known Cases of False-Positive Identifications Based on Flawed Fingerprint Analysis

The FBI’s false-positive identification of Brandon Mayfield is certainly not an isolated anomaly. In State v. McPhaul, 808 S.E.2d 294 (N.C. Ct. App. 2017), Juan Foronte McPhaul was convicted by a jury of numerous felonies, including attempted first-degree murder, assault with a deadly weapon with intent to kill inflicting serious injury (“AWDWIKISI”), and robbery with a dangerous weapon for his role in the robbery and beating of Domino’s Pizza driver Tyler Lloyd. The trial court imposed consecutive terms of 238-298 months for attempted murder, 88-118 months for AWDWIKISI, and 97-129 months for the robbery.

On appeal, McPhaul argued, among other things, that the trial court erred by allowing the State’s expert witness, Trudy Wood, to testify that latent fingerprints found on Lloyd’s truck and on items taken during the robbery but later seized from McPhaul’s residence matched McPhaul’s fingerprint records.

The Court observed that “[p]ursuant to amended N.C. Gen. Stat. section 8C-1, Rule 702(a), (a), if scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion, or otherwise, if all of the following apply:

The testimony is based upon sufficient facts or data.

The testimony is the product of reliable principles and methods.

The witness has applied the principles and methods reliably to the facts of the case.”

In McPhaul’s case, Wood’s testimony failed to satisfy criterion (3). The Court observed that she “previously testified that during an examination, she compares the pattern type and minutia points of latent print and known impressions until she is satisfied there are ‘sufficient characteristics and sequence of similarities’ to conclude that the prints match. However, Wood provided no such detail in testifying how she arrived at her actual conclusions in this case. Without further explanation for her conclusions, Wood implicitly asked the jury to accept her expert opinion that the prints matched. Since Wood failed to demonstrate that she ‘applied the principles and methods reliably to the facts of the case,’ as required by Rule 702(a)(3), we hold that the trial court abused its discretion by admitting this testimony.” However, the Court determined that, in light of the State’s other evidence against McPhaul, the error was harmless.

Another case is that of Dwight Gomas, who spent 17 months in jail in New York awaiting trial for a robbery he did not commit. Detectives performing a routine review of the evidence discovered that his fingerprints did not match crime scene evidence. Gomas, a native of Georgia, was more than 800 miles away in Atlanta when the crime occurred. He was set free after the error was uncovered and settled a lawsuit against the city for $145,000.

On May 30, 1997, an officer with the Boston Police Department (“BPD”) was shot with his own service weapon after a struggle with an unknown assailant. The assailant fired an additional shot at a person standing in the window of a second-floor bedroom. The assailant then fled the scene, leaving behind a baseball cap he had been wearing. He forcibly entered a nearby home and stopped to drink from a glass of water. He then fled the home, leaving behind the gun and a sweatshirt he had been wearing.

Two weeks later, the BPD officer identified Stephen Cowans from an eight-photo array. On July 2, 1997, the officer singled out Cowans in a lineup as the man who shot him. The person who had been watching from the second-floor bedroom window also identified Cowans as the perpetrator. At trial, the prosecutor also presented testimony from an expert examiner that a latent print recovered from the drinking glass used by the assailant matched Cowans’ left thumbprint. Cowans was convicted.

Then, on May 22, 2003, the Suffolk Superior Court approved a stipulation agreement between the Innocence Project, Cowans’ counsel, and the Commonwealth for the release of the drinking glass, the baseball cap, and the sweatshirt for the purposes of DNA testing. Those tests revealed that the primary DNA profile from the cap matched the profile taken from the drinking glass; however, Cowans was excluded as the source of either profile.

In January 2004, DNA testing of the sweatshirt revealed a DNA profile matching that from the cap and drinking glass, but again, Cowans was excluded as the source. The Suffolk County Attorney then had the fingerprint used to convict Cowans reanalyzed and concluded the print did not belong to Cowans.

On January 23, 2004, Cowans walked out of prison a free man, after having been imprisoned for five and a half years for a crime he did not commit. On February 2, 2004, he was officially exonerated—becoming the 141st person exonerated as a result of postconviction DNA testing. Sadly, on October 26, 2007, Cowans was found dead in his home from gunshot wounds.

“Our criminal justice system is generally slow to respond to any kind of science-based innovation.” Tom Albright, neuroscientist at the Salk Institute for Biological Studies in La Jolla, California

Progressive Steps Forward in Ensuring Accurate Fingerprint Analysis and Identifications

The NAS Report made clear that the ACE-V method for Analysis, Comparison, Evaluation, and Verification is “not specific enough to qualify as a validated method for this type of analysis.” Merely following the steps of that broadly stated framework “does not imply that one is proceeding in a scientific manner or producing reliable results.”

In addition, the NAS Report asserted that error rates exist, and none of the variables that fingerprint examiners rely upon have been “characterized, quantified, or compared.” Absent any statistical data, fingerprint examiners are relying on “common sense” or “intuitive knowledge,” but not on “validated information or research.”

The 2016 PCAST report concluded that while “foundationally valid,” fingerprint analysis “should never be presented in court without evidence of its error rates and of the proficiency or reliability of not just the method, but the particular examiner using the method.”

Yet for more than 100 years, fingerprint results were treated by the forensic science community, the courts, and jurors as infallible—or nearly so. In 1985, the FBI’s manual The Science of Fingerprints: Classification and Uses declared, “Of all the methods of identification, fingerprinting alone has proved to be both infallible and indefeasible.” Even as late as 2003, the head of the FBI’s fingerprint unit said in a segment that aired on the TV news program 60 Minutes that “the probability of error in fingerprint analysis is 0 percent, and that all analysts are and should be 100 percent certain of the identifications that they offer in court.”

Such hyperbole is both unscientific and unsustainable. Ironically, shortly after the news program aired, the FBI was forced to admit that at least three of its very best fingerprint examiners made the false-positive identification of Brandon Mayfield.

Improving the accuracy and reliability of fingerprint evidence can perhaps be achieved by a combination of limiting the language fingerprint experts use when testifying as to their opinions, better recovery of patent and latent prints from crime scenes, and better analysis during the comparison process.

A 2017 report from the American Association for the Advancement of Science (“AAAS”) added that fingerprint examiners should eliminate statements that contribute to the “misconceptions” shared by members of the public due to “decades of overstatement by latent print examiners.” Specifically, the AAAS asserted that terms like “match,” “identification,” “individualization,” and other labels should not be used by examiners, nor should they make any conclusions that “claim or imply” that only a “single person” could be the source of a print. Instead, latent fingerprint examiners should, at most, “state that they observe similarity between a latent print and a known print, and that a donor cannot be excluded as the source.”

Improving analysis and comparison would perhaps logically follow the removal of human judgment from the method. Beginning on January 24, 2024, the National Institute of Standards and Technology (“NIST”) tested biometric algorithms for latent fingerprint matching using Hisign, Dermalog, ROC, Neurotechnology, Griaule, Idemia, Peking University, and Innovatrics for its Evaluation of Latent Fingerprint Technologies (“ELFT”). Idemia and Hisign scored the highest success rate, at 96.5%. Idemia also scored the lowest false-positive identification rate (“FPIR”), at a little less than 8%. Hisign’s 11% FPIR was second. As for speed, ROC had the fastest mean mated search duration at 15 seconds—over 100 times faster than the second-fastest algorithm.

Idemia announced it was the fourth consecutive time that it had “achieved the best accuracy and the top speed among the most accurate algorithms.” Idemia had also scored high marks in fingerprint biometric accuracy during a study in late 2023.

Newer methods of recovering prints are being researched. Crime scenes are generally messy. Therefore, it’s not surprising that investigators often collect overlapping and weak fingerprints rather than perfectly placed whorls and swirls. And traditional photography cannot separate overlapping prints, leaving the evidence virtually useless.

However, in a recent study by researchers at Aarhus University, they demonstrated that it is possible to use desorption electrospray ionization mass spectrometry (“DESI-MS”) to analyze overlapping and weak fingerprints lifted using gelatin lifters.

DESI-MS works by measuring the chemical compounds in these collected fingerprints based on their mass. Invented 20 years ago for general surface analysis, DESI-MS was shown in 2008 to be viable for chemical imaging of fingerprints on glass surfaces and tape.

The researchers are also working to employ the chemical imaging method for profiling. Since fingerprints can reveal much more than a pattern—including the use of nicotine, drugs, and caffeine, as well as potentially incriminating substances such as lubricant from condoms—recovering these compounds provides additional evidence. Other researchers across the globe are also working on this. It’s possible to utilize a variety of methods to collect the samples and even gather data concerning individuals’ gender, age, and lifestyle, which would be valuable identification evidence.

Another means of improving the reliability of fingerprint evidence is by learning from previous mistakes. The Office of the Inspector General (“OIG”) investigated the causes of the misidentification of Mayfield. Some of the OIG’s conclusions were:

Although Mayfield and the true source of the prints, Ouhnane Daoud, did not have identical prints, they nevertheless had very similar-looking prints;

After FBI examiners found 10 points of unusual similarity between Mayfield’s fingerprint and the latent fingerprint, “the FBI examiners began to find additional features in [the latent print] that were not really there, but rather were suggested to the examiners by features in the Mayfield prints. As a result of this process, murky or ambiguous details in [the latent print] were erroneously identified as points of similarity with Mayfield’s prints”; and

Although the Spanish National Police advised the FBI on April 13, 2004, that the latent fingerprint recovered from the bag of detonators did not match Mayfield’s prints, the FBI nevertheless arrested Mayfield more than three weeks later, on May 6, 2004, in what the OIG concluded was the result of “the FBI laboratory’s overconfidence in the skill and superiority of its examiners [which] prevented it from taking the [April 13 report] as seriously as it should have.”

The results of an experiment appearing in Cognitive Research Journal (2020) also shed light on methods designed to reduce the risk of fingerprint misidentification. In the experiment, the researchers extended the “wisdom of the crowds” approach to fingerprint analysis, comparing the performance of individuals and crowds of professional fingerprint analysts. Both crowds of novices and crowds of seasoned experts were tested, and their responses were aggregated using three different rules:

“Follow the majority”—meaning the judgment with the most group support was adopted;

“Follow the most confident”—meaning the judgment with the highest confidence rating was adopted;

“Follow the most senior”—meaning the judgment of the most experienced examiner was adopted.

Thirty-six professional fingerprint examiners from the Australian Federal Police volunteered to be the group of seasoned experts. For the novice group, 36 undergraduate psychology students from the University of Queensland and the University of Adelaide volunteered.

Volunteer undergraduate students left their prints on wood, plastic, metal, and glass surfaces, and these simulated “crime scene prints” were dusted, photographed, cropped, and isolated in the frame. Each simulated print was certified by an expert to contain sufficient information to make an identification as to whether or not there existed a clear comparison exemplar.

Each of the fingerprint examiners was presented with the same set of 24 fingerprint pairs from the same finger (targets) and 24 highly similar pairs from different fingers (distractors) in differing random orders. Each of the pairs consisted of a “latent” crime scene print and a fully rolled “arrest” fingerprint. Each examiner was given 48 trials, and examiners had only 20 seconds to examine 44 pairs but an unlimited amount of time to examine the final four pairs.

In the 20-second condition, individual experts correctly made true-positive identifications 71% of the time but incorrectly made false-positive identifications 8.5% of the time. By comparison, individual novices also made true-positive identifications 71% of the time but made false-positive identifications 50% of the time.

In the four untimed comparisons, individual experts made true-positive identifications 85% of the time and false-positive identifications 2.8% of the time. However, individual novices made true-positive identifications 76% of the time and false-positive identifications 60% of the time. (These results suggest that novices are prone to conclude prints are a “match.”)

After tallying the results according to the three methods stated above, it became obvious that “follow the majority” provided the most accurate conclusions for the experts. For example, in the four untimed comparisons, the correct true-positive identifications rose from 85% for individual examiners to 96% for groups of three experts. Similarly, the incorrect false-positive identifications fell from 2.8% for individual experts to 0% for groups of three experts.

However, for novices, the correct true-positive identifications for individual examiners rose only to 79% for novice trios, and the 60% rate of incorrect false-positive identifications for novices actually rose to 79% for groups of three novices.

While the saying “two heads are better than one” (or in this experiment, three heads) reflects common sense when tackling difficult problems, in the case of friction ridge analysis, it must be three well-trained and experienced heads.

Finally, in seeking to improve the current method of fingerprint comparison and analysis, an expert working group convened by the NIST and the National Institute of Justice recommended the creation of “a culture in which both management and staff understand that openness about errors is not necessarily a path to punitive sanctions but rather is part of an effective system to detect deviations from desired practices and incorrect judgments in latent print casework.” Recommendations for management in forensic laboratories included:

Employ a system to identify and track errors and their causes.

Establish policies and procedures for case review, conflict resolution, corrective action, and preventive measures.

The working group also generated in its report a “flow chart that shows the Analysis, Comparison, Evaluation, and Verification (ACE-V) process for latent print examination that is currently used in the nation’s forensic crime laboratories.” The flow chart “offers a detailed view of steps in the ACE-V process where human error risks should be minimized.”

Furthermore, the group identified a “critical need” for research into the interpretive process that is at the core of the ACE-V process, to include: an analysis of educational and cognitive abilities that could help with successful analysis; the creation of large “test” databases to test trainees and train AI systems; and a formulation of the most effective training methodologies and tests to improve examiner performance and boost success rates. The field also needs better objective metrics to assist in matching latent prints at all stages of the process.

There is a huge opportunity for improvement in automated and AI testing platforms. These need further development and training and should be evaluated objectively and scientifically. Far from being unassailable or perfect, this is a forensic discipline in the process of development and improvement.

Conclusion

The case of Brandon Mayfield and numerous other documented false-positive identifications expose the troubling flaws in fingerprint analysis, a forensic method long (mistakenly) regarded as infallible. Despite its widespread acceptance in courts, fingerprint comparison remains highly subjective, lacking standardized thresholds for declaring a match and suffering from cognitive biases that can lead to wrongful convictions. Studies reveal alarming error rates, and the ACE-V method—though widely used—has not been scientifically validated to ensure consistent, reliable results.

Courts, often deferring to precedent rather than rigorous scientific scrutiny, continue admitting fingerprint evidence under the assumption that cross-examination will expose its weaknesses, yet this reliance on adversarial testing has repeatedly proven inadequate. The legal system’s commitment to finality further entrenches these flaws, leaving little recourse for those wrongly convicted based on faulty fingerprint evidence.

Clearly, improvements are needed and must include stricter expert testimony guidelines, enhanced training to mitigate bias, and greater integration of automated systems to reduce human error. Until fingerprint analysis is subjected to the same scientific rigor as DNA testing, its reliability will remain questionable—and the risk of injustice will persist—despite being viewed by most as virtually infallible.

The path to reform requires a fundamental shift in how forensic evidence is evaluated—both in labs and courtrooms. Blind verification, algorithmic assistance, and transparency in error reporting are essential steps toward ensuring accuracy. The legal system must abandon its uncritical acceptance of fingerprint evidence and instead demand empirical validation, standardized protocols, and acknowledgment of its limitations. Only then can fingerprint analysis evolve from a historically flawed practice into a truly reliable forensic tool that aids the pursuit of justice, rather than undermining it.  

Sources: nikouiandassociates.com, bu.edu, ojp.gov, sciencenews.org, open.mitchellhamline.edu, direct.mit.edu, cognitiveresearchjournal.springeropen, theconversation.com, biometricupdate.com, forensicmag.com, uclalawreview.org, archives.fbi.gov, justiceflorida.com, innocenceproject.org

Subscribe today

Already a subscriber? Login