Return to Transcripts main page

CNN Presents

Encore Presentation: Reasonable Doubt

Aired May 07, 2006 - 14:00   ET

THIS IS A RUSH TRANSCRIPT. THIS COPY MAY NOT BE IN ITS FINAL FORM AND MAY BE UPDATED.

(NEWS BREAK)
UNIDENTIFIED FEMALE: That was the worst moment of my life.

ANNOUNCER: Her son sent to prison for life for a murder he didn't commit.

UNIDENTIFIED FEMALE: It was like someone had hit me in the stomach. I couldn't believe it.

ANNOUNCER: Her brother convicted of murder may be innocent.

UNIDENTIFIED FEMALE: The bullet evidence was so shocking and devastating when the FBI came and testified.

ANNOUNCER: Fingerprints. Matching bullets. The cases were airtight. The forensic evidence unassailable. Or was it?

UNIDENTIFIED MALE: These people could put my sons in jail with bogus science.

UNIDENTIFIED FEMALE: There are no requirements and no standards that must be met.

UNIDENTIFIED MALE: You can take a two week course and declare yourself an expert and get hired and make life and death decisions.

RIKY JACKSON, FRIEND MURDERED: My friend was beheaded practically. He was multiply stabbed.

ANNOUNCER: Riky Jackson remembers the day his buddy Alvin Davis was murdered.

JACKSON: A mutual friend of Alvin's and mine lived across the street and I'll never forget, he called me, he said something's happening in your friend's building. My involvement came when I instructed him, well, if the need, if there is any assistance I can lend, give them my name and number. Silly me.

ANNOUNCER: A few weeks later the police called him down to the station in Upper Darby Township, near Philadelphia. They had found his bloody fingerprints, they said, in the victim's apartment. They arrested him for murder.

RICHARD JACKSON, RIKY JACKSON'S FATHER: They showed me vivid, color pictures of fingerprints and they were confident that these prints were Riky's. ANNOUNCER: Richard and Verna Jackson refused to believe their son could be a vicious killer, so they hired defense attorney Mike Malloy. With fingerprint evidence, he knew what he was up against.

MIKE MALLOY, DEFENSE ATTORNEY: Jurors believe that a fingerprint is almost a fait accompli for guilt. That those prints are his is really saying that he was at the crime scene.

ANNOUNCER: Malloy hired his own expert, a retired FBI examiner.

MALLOY: He says to me in his southern drawl, Mike, those prints do not match. As soon as he put his eye to the magnifying glass he knew instantly.

ANNOUNCER: The lawyer then got a second opinion. Another retired FBI examiner with the same results. No match.

But at the trial, three prosecution examiners testified that the bloody fingerprints matched Riky Jackson's prints without any doubt. It only took the jury a few hours to reach a verdict, guilty of first degree murder.

VERNA JACKSON, RIKY JACKSON'S MOTHER: That was the worst moment of my life. I felt like I was dying that day.

ANNOUNCER: Jackson began serving a life sentence with no chance of parole.

JACKSON: As far as being in jail. Holidays suck, your mother's birthday is horrendous.

I found basically it was a matter of not wanting to wake up, oh, but damn I am.

ANNOUNCER: Meanwhile, Malloy's experts were so outraged by the verdict that they sent the prints to be reviewed by a group that accredits fingerprint examiners. Their verdict? The prints didn't match Jackson's. Even though the case was closed, Malloy kept pushing.

MALLOY: About a year or so after the conviction, after maybe what would be the 20th motion we filed, the prints were eventually sent to the FBI. And they came back, again, quickly said, no, they don't match.

ANNOUNCER: The fingerprint evidence that put Riky Jackson in prison for life was simply wrong. After more than two years in prison, Jackson was released.

VERNA JACKSON: Our lawyer met us at the door and he said we're bringing Riky home today. And you would never understand this unless you went through it and, sorry to be breaking up - It was two days before Christmas.

ANNOUNCER: How could fingerprints be wrong? The FBI text on fingerprint identification claims, "Fingerprinting has proved to be infallible." So how could three different examiners all make the same mistake that sent an innocent man to prison.

From the way it's presented on television shows like "CSI" or in courtrooms, you would think forensic evidence like fingerprints, hair samples, bullet analysis, is infallible. High tech detectives nailing the culprit with scientific precision. But it turns out the aura of infallibility is a myth. Forensic evidence is subject to human error. Crime labs are unregulated. Standards are ad hoc or even non-existent and fingerprint identification, the old standard of forensic evidence, is more art than science.

SIMON COLE, UC IRVINE: I think the analogy with an art connoisseur is not all that wrong. Somebody who has spent a lot of time studying art and thinks this is a Rembrandt and says it's a Rembrandt, and that's it, it's just their opinion that it's a Rembrandt. We can't verify it scientifically in any way.

ANNOUNCER: Fingerprint identification has been used in courts for almost 100 years. Even critics say examiners probably get it right most of the time. The problem is juries are led to believe it is accurate 100 percent of the time and surprisingly, the reliability of fingerprint identification has never been tested.

COLE: There hasn't been a validation study designed to measure the accuracy of fingerprint identification.

ANNOUNCER: Simon Cole, a nationally recognized expert on fingerprint evidence, showed us how matching is done.

COLE: What a fingerprint examiner would do in making a comparison is find occurrences along the ridges on the unknown print and see if those are also present in the known print.

ANNOUNCER: There is no generally accepted standard for how many points of similarity make a match, even though such conclusions are reached in courtrooms on a regular basis.

COLE: One law enforcement agency would have eight points. One would have ten points. One would have twelve points. So now we have a new system in which we don't need any number at all, it's up to the examiner to decide how much matching consistency they need between the known and unknown print.

ANNOUNCER: And there is evidence plenty of mistakes get made. The International Association for Identification tests fingerprint examiners who want to be certified. Nearly half of those who take the test fail it and those are the ones that take the trouble to get certified.

COLE: But you don't need to be certified to testify in court. There are thousands of uncertified examiners testifying in this country.

ANNOUNCER: Still, experts present their findings as indisputable. Take this affidavit about a fingerprint where an FBI agent states, "The FBI lab stands by their conclusion of 100 percent positive identification." That fingerprint was on a bag containing detonators and explosives found near the Madrid train bombing attacks that killed nearly 200 people. FBI examiners determined the print on the bag belonged to an Oregon attorney, Brandon Mayfield.

COLE: The prints were also looked at by a fingerprint examiner retained by Mayfield himself and that examiner also agreed that the print belonged to Mayfield.

ANNOUNCER: But Spanish officials matched the print to an Algerian national living in Spain. Brandon Mayfield was released after spending two weeks in prison.

BRANDON MAYFIELD, VICTIM OF FINGERPRINT ERROR: I'll just say that it was humiliating, it was embarrassing.

BOB JORDAN, FBI SPECIAL AGENT: The FBI regrets the hardships that this matter has placed upon Mr. Mayfield and his family.

ANNOUNCER: The FBI declined to talk to CNN about the Mayfield mistake. About how 100 percent certainty could be 100 percent wrong.

But an investigative panel found the culture at the Bureau compounded the original human error, noting, "Once the mindset occurred with the initial examiner, it became increasingly difficult for others in the agency to disagree."

Such mistakes may be rare, but no one knows how many more Riky Jacksons are out there, how often supposedly infallible forensic science helps convict innocent people while allowing the guilty to go free.

Next on CNN PRESENTS, murderer or victim of flawed forensics?

UNIDENTIFIED MALE: Neither assumption that the FBI was using was warranted.

ANNOUNCER: Decades of FBI testimony in doubt.

(COMMERCIAL BREAK)

ANNOUNCER: Jacqueline Behn shared a passion with her baby brother Michael, collecting coins.

JACQUELINE BEHN, MICHAEL BEHN'S SISTER: I think I was the one who gave him his first coin.

ANNOUNCER: It was quite a shock for her when his collecting got him arrested for armed robbery and murder.

BANE: It was like someone had hit me in the stomach. I couldn't believe it. He had never been in trouble.

ANNOUNCER: The murder took place here on Main Street in South River, New Jersey. Michael Behn, who had no previous criminal record, was accused of the execution-style shooting of a coin dealer in order to steal $40,000 worth of rare coins. Police found the coins at the home of Behn's mother. He told them he bought the coins from the murdered coin dealer, Robert Rose, the day he was killed. CNN interviewed Michael Behn inside a New Jersey state prison.

UNIDENTIFIED MALE: How well did you know Robert Rose.

MICHAEL BANE, ACCUSED OF MURDER: Well enough for him to know my name. His wife knew me. I had seen him at coin shows before.

ANNOUNCER: Bane says he and Rose did an under-the-table cash deal. No paperwork, no sales receipt. No proof he had bought the coins.

MICHAEL BEHN: That's commonplace in the coin business. It's like the diamond business. There's a lot of cash.

ANNOUNCER: At trial, prosecutors produced no eyewitness, no fingerprint, no murder weapon. But an FBI examiner testified that bullets found in Michael Behn's office and the bullets that had killed the coin dealer, quote, "came from the same source of lead at the manufacturer, so they were manufactured on our about the same date."

JACQUELINE BEHN: The bullet evidence was so devastating and shocking when the FBI came and testified that day that I couldn't believe. When you are told that bullets could potentially match, there's nothing worse than that.

ANNOUNCER: Jacqueline's brother was convicted of first degree murder and armed robbery.

In handing down a sentence of 30 years to life, the judge called the bullet evidence "particularly significant." It was the only physical evidence linking Behn directly to the killing.

The process is called "comparative bullet lead analysis" and the FBI is the only lab in the country that performs it. By testing bullets for tiny amounts of impurities, FBI examiners find a bullet's chemical profile. If the crime scene bullets and the suspect's bullets turn out to have identical profiles, it's a virtual match. The FBI has claimed in court to be able to link one bullet to others from the same production run, even from the same box.

Bullet analysis has been used in thousands of criminal cases as far back as the investigation of the assassination of President Kennedy. The working assumption is that each batch of bullets unique, but it just didn't make sense to Jacqueline Behn.

JACQUELINE BEHN: I just intuitively thought that it would be virtually impossible with billions and billions of bullets being manufactured every year that they could pinpoint it in this way.

ANNOUNCER: Behn, who happens to be an associate professor of criminology, went on a mission to research the science of bullet lead analysis.

JACQUELINE BEHN: And I was very surprised to note that the peer review didn't really exist, that the well-documented journal research didn't exist for this topic. Very surprised.

ANNOUNCER: The few publications she found only raised more questions. Was it possible that there was no scientific foundation for the testimony that had convicted her brother?

Behn needed somebody to test the FBI's science. Metallurgist Eric Randich was more than willing.

ERIC RANDICH, METALLURGIST: The reason I took on this scientific study is that I recognized that these people could put my sons in jail with bogus science.

ANNOUNCER: Randich contacted two lead smelters that supply ammunition manufacturers.

RANDICH: Each of these rows is one lot of lead. That is, one source, as the FBI would define it, of lead. Typically, each of these lots, for example, 15 million .22 caliber bullets.

ANNOUNCER: He questioned two fundamental assumptions the FBI made about lead. That every batch has a perfectly consistent composition and that every batch of lead is unique.

RANDICH: We found that, in fact, neither assumption that the FBI was using was warranted.

ANNOUNCER: His research found different batches of lead, some produced more than a decade apart, that were identical according to the FBI's standard. This undercut decades of FBI testimony in cases like Michael Behn's. Matches found by the FBI could be meaningless.

RANDICH: You can say that the bullets are indistinguishable, but you don't know how likely it is they came from the same box. Or boxes made the same day. We just do not have enough information to know the probability.

ANNOUNCER: Before publishing his findings, Randich approached the FBI, whom he had worked with on previous occasions.

RANDICH: And I said, would you like to look at the data? And they said, no, we just have a difference of opinion.

ANNOUNCER: FBI experts continue to provide court testimony counter to Randich's research. For example, the FBI examiner who had testified in Michael Behn's trial told an Alaska jury, "We have looked at different sources of lead and never found two to be exactly the same."

But more research continued to raise questions. Defense attorneys began a legal assault on bullet lead analysis. At the time, the FBI issues a statement saying it has been using these techniques for 20 years and is "very confident that all of our methods are fully supported by scientific data."

However, the FBI commissioned a study by the National Academy of Sciences. The Academy's findings, issued last year, were stunning. They cast serious doubts on the central premise of bullet lead analysis and appeared to contradict nearly 40 years of FBI testimony.

"The available data do not support any statement that a crime bullet came from or is likely to come from a particular box of ammunition."

In response, the FBI stood by the usefulness of bullet lead analysis, which it said had "continually withstood legal challenges in hundreds of cases."

Bullet lead analysis is just one of many forensic techniques, like fingerprinting and hair analysis that have recently come under suspicion. Experts tell CNN such evidence has actually evolved through intuition, anecdote and human trial and error.

DAVID FAIGMAN, UC HASTINGS COLLEGE OF LAW: You have this huge mass of forensic identification technology that really has little or no empirical support underlying it.

And I'm not sure how they can testify with such certainty.

ANNOUNCER: The FBI declined to discuss bullet lead analysis or Michael Behn's case with CNN. This past November, a U.S. District Court judge refused to allow the same FBI examiner who testified against Behn to present bullet lead analysis in another case, finding the FBI did not offer "sufficient scientific basis" for the analysis.

Michael Behn, imprisoned for almost 10 years, maintains his innocence and has petitioned for a new trial.

MICHAEL BEHN: If you look at the evidence, it shows that I didn't do it.

JACQUELINE BEHN: I have faith in the system and I really believe that they will pay attention.

ANNOUNCER: When CNN PRESENTS returns.

UNIDENTIFIED MALE: My discharge date was 2025 and I didn't really plan on ever getting out of prison.

ANNOUNCER: Three overturned convictions, one thing in common, the forensic scientist.

(COMMERCIAL BREAK)

UNIDENTIFIED FEMALE: This is my first time as a tattoo artist.

ANNOUNCER: Jimmy Ray Bromgard's fiancee is putting the finishing touch on a one-of-a-kind reminder of his life behind bars.

JIMMY RAY BROMGARD, : All the years I was in prison. Starting with June, 1987 and then '88. '89, '90, '91, '92 and Lashaw (ph) is going to put October 1st, 2002 on there, which is the day that our old lives ended and our new lives started. When I got released, so ...

ANNOUNCER: It was an end of a fifteen year nightmare that began with a police lineup in Billings, Montana.

At the time, 19 year old Bromgard was serving 20 days for a misdemeanor, fighting at school.

BROMGARD: And they asked me if I wanted to be in a lineup, I said, "Sure." Gets me out of the cell. They didn't tell me why, they just gave us a piece of paper and we had to read these lines that said, "Shut up or else I'll kill you."

ANNOUNCER: An eight year old girl picked out Bromgard as the man who assaulted her. He was charged with rape. At trial the child testified she was unsure Bromgard was her attacker but apparently the forensic evidence was compelling.

BROMGARD: This is the third day of my trial. And the jurors, they said that several of the jurors were convinced I was guilty the second day of my trial and it was a three-day trial.

ANNOUNCER: The second day was the day Arnold Melnikoff testified.

Melnikoff was the founder and manager of Montana's state crime laboratory. He testified that hairs taken from Bromgard and hairs from the crime scene were "almost indistinguishable" and "microscopically similar." He also told the jury there was little chance, based on statistics, that the crime scene hair came from anyone other than Bromgard.

BROMGARD: He gave these numbers that it was a 1 in 10,000 chance that it wasn't me that did it. He said, more or less he told the jury this had to be the guy that did it.

ANNOUNCER: The jury took about an hour to reach a guilty verdict. Bromgard was handed three concurrent 40 year sentences and shipped off to state prison.

BROMGARD: And they put me out in general population where I got my jaw broke for being a child molester.

ANNOUNCER: Bromgard spent a lot of time in lockdown, in his cell 23 hours a day. BROMGARD: The first time was for like nine months, then 12 months, then 15 months. Something like that.

ANNOUNCER: He did file appeals, but to no avail.

BROMGARD: My discharge date was 2025 and I didn't really ever plan on getting out of prison.

ANNOUNCER: Then, in 2000, he saw a story on TV about the Innocence Project, a group of lawyers who got convictions overturned by retesting evidence. Bromgard asked his attorney to contact them.

PETER NEUFELD, ATTORNEY, THE INNOCENT PROJECT: We took on the case and we filed a motion in court in Billings, Montana and a judge ordered post-conviction DNA testing. ANNOUNCER: The DNA results were startling. They proved Bromgard was innocent.

UNIDENTIFIED MALE: And is released forthwith.

ANNOUNCER: Bromgard was free but Peter Neufeld wanted to know what had gone wrong. He asked a panel of experts to review Melikoff's testimony. One of the experts was Harold Deadman, a former FBI hair examiner who teaches forensics at George Washington University.

Deadman told CNN that numerous mistakes in hair analysis have been uncovered around the country.

HAROLD DEADMAN, GEORGE WASHINGTON UNIVERSITY: There are many laboratories where individuals are put into the position of doing hair analysis with very little training.

ANNOUNCER: Surprisingly, hair examiners, like fingerprint examiners, are largely unregulated. In all but a few states, all it takes to testify is a judge's OK. In the Jimmy Ray Bromgard case, the experts' assessment of Melnikoff's testimony was blunt.

DEADMAN: Egregious misstatements, not only of the science of forensic examinations, but also of genetics and statistics. ANNOUNCER: As for Melnikoff's statistic of a 1 in 10,000 chance the hairs were not Bromgard's the experts say, "There was not and never was a scientific basis to make any such calculation."

Melnikoff declined to be interviewed but his attorney, Rocky Treppiedi, sat down with CNN. He says his client now admits his calculations were off.

ROCKY TREPPIEDI, MELNIKOFF'S ATTORNEY: That may not be the exact mathematical approach that one should take to reach that probability.

ANNOUNCER: Neufeld wanted to know if the problem was more than bad statistics. He convinced the district attorney to send the actual hairs to FBI experts. The concluded the crime scene hairs could not possibly have come from Bromgard. NEUFELD: They said Melnikoff got it all wrong. In fact, the FBI says the head hair looks like it could have come from the little girl.

MELNIKOFF: According to his own testimony, Melnikoff had evaluated hair in as many as 700 investigations. Neufeld, worried there might be other wrongful convictions, did a computer search and came across two cases. In one case, DNA tests had resulted in the conviction overturned just like Bromgard's. In the other case, DNA tests had not been done. Neufeld contacted the convicted man and had the 14-year-old evidence tested.

NEUFELD: We did the DNA testing, sure enough, he was excluded and exonerated as well.

ANNOUNCER: Neufeld also contacted the attorney general's office in Washington State, where Melnikoff worked after leaving Montana. The state police had experts review a sample of Melnikoff's work. Their conclusion: though his results were correct, his methods were badly flawed. The report charged him with neglect of duty, incompetence, gross misconduct and violation of agency rules. Melnikoff's attorney says the state's review was conducted improperly.

(BEGIN VIDEO CLIP)

UNIDENTIFIED MALE: Those are bogus findings quite simply. They don't have anything to pin that on. That's certainly what their report says, but that isn't what the facts say.

(END VIDEO CLIP)

UNIDENTIFIED MALE: Washington state patrol officials tell CNN they stand by their report. Based on their finding that his testimony in Montana was erroneous, they dismissed Melnikoff in 2004. He's appealing his dismissal.

CAPT. FRED FAKKEMA, WASHINGTON STATE PATROL: His credibility, not only to the scientific community, but also with inside the justice system here, and for the Washington state patrol, was in jeopardy and we felt that this was the appropriate measure to take.

UNIDENTIFIED MALE: The shoddy work of one forensic scientist can affect hundreds, even thousands of cases. Melnikoff had been a forensic scientist for 30 years before anyone started questioning his methods and results and his long career is not unique.

UNIDENTIFIED FEMALE: I devoted my entire career to that job.

UNIDENTIFIED MALE: Joyce Gilcrist (ph) collected and analyzed evidence for the Oklahoma City Police for 21 years before her mistakes were caught. After a judge criticized her testimony, FBI and internal police reviews turned up problem after problem with Gilcrist's work. The misidentification of hairs helped send Jeffrey Pearce (ph) to prison. Pearce is one of at least three men whose convictions have been overturned. One of them was on death row.

Gilcrist was fired in 2001. Now hundreds of cases spanning two decades have to be reviewed. Gilcrist declined to speak with CNN, but her attorney maintained she did nothing wrong.

UNIDENTIFIED MALE: No comment.

UNIDENTIFIED MALE: In West Virginia, Fred Zane (ph) got away with 12 years of sloppy work and false testimony in a crime lab before DNA testing caught up with him. The state invalidated his entire body of work.

Zane's work helped put at least six men in prison erroneously in two states. He was indicted for false testimony, but died before he could be brought to trial. There's no way of knowing how many other Fred Zanes are out there.

UNIDENTIFIED MALE: The way we assess professionalism is by having rigorous proficiency testing, quality assurance programs, meaningful audits. We don't have any of those in place. UNIDENTIFIED MALE: You take a two-week course and call yourself an expert and get hired and make life and death decisions.

UNIDENTIFIED MALE: Coming up, a crime lab shut down.

(BEGIN VIDEO CLIP)

BILL THOMPSON: I was frankly appalled at what I saw.

(END VIDEO CLIP)

(COMMERCIAL BREAK)

UNIDENTIFIED MALE: DNA profiling is the gem of forensic evidence. It is a powerful tool. Unlike hair analysis or fingerprint identification, it's steeped in science. There are even national standards.

UNIDENTIFIED MALE: DNA profiling arose out of the scientific community. It's been more or less supervised by the scientific community.

UNIDENTIFIED MALE: But even DNA testing is not infallible. Results can be tainted by misinterpreting. Nowhere has this been of greater concern than in Houston, Texas.

THOMPSON: The problems with the Houston DNA lab, the list is lengthy.

UNIDENTIFIED MALE: In 2002, Houston area journalists asked Professor William Thompson, an expert in scientific evidence, to review a handful of lab reports from the DNA unit of the Houston police department's crime lab. The reporters had been tipped off by defense attorneys who had become suspicious about the lab reports' accuracy.

THOMPSON: I was frankly appalled at what I saw. They were engaging in certain practices that were dangerous and conducive to error. They were presenting the results to juries in ways that in some cases were outright wrong.

UNIDENTIFIED MALE: The lab commissioned an outside review, which confirmed many of Thompson's criticisms. The DNA lab was shut down, but not soon enough for Josiah Sutton (ph). Sutton was a high school kid when he was arrested and convicted of rape after an examiner from the Houston crime lab identified his DNA in the evidence. In the midst of the widening lab scandal, Thompson was asked by reporters to review the Sutton case.

THOMPSON: In that case, not only was it clear that the lab work was done poorly, but it appeared that we had a man who had been falsely incriminated.

UNIDENTIFIED MALE: The evidence was retested and Sutton's DNA was not found. After 4 1/2 years behind bars, Sutton was released. The district attorney's office had a mess on its hands. Evidence in every case in which a DNA test had played a role in the conviction would be retested. Assistant district attorney Marie Monet (ph) was put in charge of the monumental task.

MARIE MONET: We started with over 1300 police reports where there was a DNA lab number associated with them.

UNIDENTIFIED MALE: Monet's office is retesting evidence in almost 400 cases. The results so far have confirmed guilt in most, but not all of them.

THOMPSON: There are quite a number of others that are suspicious.

UNIDENTIFIED FEMALE: When we get done with all of this, there will be a certain number of cases where the laboratories cannot confirm or deny the original HPD testing.

UNIDENTIFIED MALE: How could this happen? Auditors found that examiners lacked training. Testing procedures were inadequate. Equipment hadn't been calibrated and a leaky roof could have contaminated DNA samples. Critics tell CNN the problems that plagued the Houston DNA lab appear to be especially egregious, but not unique. Bill Thompson says many other labs have reported errors.

THOMPSON: It's not uncommon to have the DNA of laboratory workers showing up in the samples being tested. It's not that uncommon to have samples mislabeled or switched and it's that uncommon to have inadvertent transfers of DNA from one sample to another sample.

UNIDENTIFIED MALE: Even the most prestigious DNA lab in the country, the FBI lab, is not immune to problems. Last year, the inspector general of the Justice Department found that FBI lab employee Jacqueline Blake (ph) failed to complete a critically important test in 90 cases. The test, which detects whether or not a DNA sample has been contaminated is called a negative control.

THOMPSON: Jacqueline Blake figured out a very clever way to make sure that she had no failures of her negative controls. She didn't run negative controls.

RALPH WARREN: No one had ever put her on notice that she was doing her job improperly.

UNIDENTIFIED MALE: Ralph Warren, Blake's attorney, talked to CNN after Blake declined to be interviewed.

WARREN: Had she known how critical the conducting of those negative controls were and how significant they were, then her reasoning is that she would have done so on each and every occasion.

UNIDENTIFIED MALE: Blake not only didn't run the tests, which experts say are standard in DNA testing, but she also falsified her laboratory documentation to conceal her improper work, according to the inspector general report. This went on for more than two years before she was finally caught purely by chance, an accidental discovery made by a colleague.

WARREN: Her direct supervisor never brought any regularities to her attention or anyone in the chain of command.

UNIDENTIFIED MALE: The inspector general found that Blake had easily escaped detection because the FBI failed to develop policies that subjected her work and the work of other DNA biologists, to adequate scrutiny. The FBI declined to discuss the Blake case with CNN.

Retests of Blake's work did not turn up any results that were wrong. She resigned from the FBI in 2002. Last year, she pled guilty in Federal court to making false statements in her lab reports. The FBI has taken steps to make it less likely another Jacqueline Blake could slip under its radar. DNA testing may be one of the best weapons in the forensic arsenal, but it must be wielded with care.

THOMPSON: Although in general it's very reliable, there can be cases where errors occur and those have to be taken into account when we start using this to make life and death decisions.

UNIDENTIFIED MALE: Next on CNN PRESENTS...

PETER NEUFELD, ATTORNEY, THE INNOCENCE PROJECT: Forensic science has gotten a free ride for the last 50 years. UNIDENTIFIED MALE: A system with little accountability.

JANINE ARVIZU, LABORATORY AUDITOR: There are not requirements and no standards that must be met.

(COMMERCIAL BREAK)

UNIDENTIFIED MALE: Riky Jackson, Jimmy Ray Bromgard, Josiah Sutton, the roll call for those wronged by forensic science. Individual tragedies but not necessarily isolated mistakes.

ARVIZU: When problems occur in forensic labs, some, the members of the public might just think that it's examples of individual malfeasance or individual problems. But from my perspective as a quality auditor, I looked at it more as a systematic failure of the forensic industry in this country. ARVIZU: Janine Arvizu has audited all kinds of laboratories over her long career. Much of what she sees in crime labs disturbs her.

ARVIZU: There are no requirements and no standards that must be met for you to be a forensic laboratory in most parts of the country.

UNIDENTIFIED MALE: In fact, in many places, all it takes to set up a crime lab is a white coat and a business card. In all but a few states, there is no oversight of crime laboratories and there is no Federal watchdog to make sure that forensic analyses, used to imprison people or send them to their death are scientific and accurate.

NEUFELD: Forensic science has gotten a free ride for the last 50 years, primarily because they made this bogus argument that we don't need to be regulated.

UNIDENTIFIED MALE: Right now, if bad lab work does come to light, nothing requires officials to determine whether it's an isolated incident or part of a pattern.

NEUFELD: There's no one to require that they conduct an audit. If after the space shuttle disaster, if NASA said, hey, mistakes happen. It doesn't work that way. In our society, every time you have a screw up like that, there are major external audits which ensue. The only institution in this country where that doesn't happen, the only institution is forensic science.

UNIDENTIFIED MALE: Neufeld has run into this problem in Montana, where crime lab director Arnold Melnikoff's faulty lab work helped convict three men whose convictions were later overturned. The state's attorney general, Mike McGrath refused to have Melnikoff's evidence scientifically retested, so the Innocence Project petitioned the state supreme court to audit all of Melnikoff's lab work but the court ruled it does not have the authority to challenge McGrath's decision.

NEUFELD: There is no doubt in my mind that if we went back, the scientists, and we examined the hair work in the hundreds of criminal cases that Melnikoff did during his tenure in Montana, there would be other wrongful convictions. There was be other exonerations. UNIDENTIFIED MALE: McGrath feels his office has met its obligations by reviewing Melnikoff's case files.

MIKE McGRATH, MONTANA ATTORNEY GENERAL: We made a decision as to whether or not his testimony made a difference in terms of the case and determined that no one else has been convicted wrongly.

UNIDENTIFIED MALE: McGrath has also refused to review Melnikoff's other forensic work from his nearly 20 years in Montana.

NEUFELD: And we haven't even cracked the surface yet on his work in arson cases, in Melnikoff's work in toxicology, drug analysis. If a man is wrong in one discipline, why wouldn't you think that he might be wrong in other disciplines as well? McGRATH: We examined every single case in the hair area, where we did have reason to check. We don't have reason to check anything else. Occasionally a sloppy performance does not mean somebody's been wrongly convicted.

UNIDENTIFIED MALE: As for Melnikoff, his attorney says he's confident the rest of his lab work has been accurate. Late last year, Congress passed legislation that could change the way officials respond to complaints about crime labs. The law expands Federal grants for forensic science, but in order to get that money, a state must establish a government entity that can conduct independent external investigations into allegations of serious negligence or misconduct at forensic laboratories. Experts applaud the legislation as a first step, but even that step is voluntary and there are still few mechanisms to catch mistakes before they get to trial in the first place. Some prosecutors see nothing wrong with the system the way it is.

McGRATH: The criminal justice system was an adversary system and the way it's designed is one person puts on evidence. The other person has an opportunity to question that evidence, to cross examine the witnesses, to do that kind of system.

UNIDENTIFIED MALE: But to critics, it's a system geared more to conviction rates than objective science. After all, many labs are headed by police officers, rather than scientists and critics say they are too close to the prosecution team.

ARVIZU: On numerous occasions, I have approached prosecutors and talked to them about the need for them to assess the quality of their own lab, not simply accepted on blind faith and almost without exception, the response from the prosecutors has been, that as long as the laboratory's results support my theory of the case, I don't want to know if there's any problems. It is hardly an appropriate response if the see no evil, speak no evil, hear no evil approach to laboratory quality.

UNIDENTIFIED MALE: When CNN PRESENTS returns...

(BEGIN VIDEO CLIP)

NEUFELD: And they were handling all this blood and it splashed around the laboratory and (INAUDIBLE) wearing gloves.

(END VIDEO CLIP)

UNIDENTIFIED MALE: Cleaning up crime labs.

(COMMERCIAL BREAK)

ROBERT SHALLER (ph): We have about 3,000 cases a year and we've performed DNA testing on all of the homicides and rapes in the city of New York.

UNIDENTIFIED MALE: Robert Shaller runs the New York City medical examiner's DNA laboratory. SHALLER: This laboratory has been responsible for making all the identifications done by DNA for the World Trade Center.

UNIDENTIFIED MALE: The lab has a stellar reputation, even among the toughest critics of crime labs, though that has not always been the case.

NEUFELD: When I first went into the New York ME's office to do an evaluation in the 1980s - remember, this is during the outbreak of HIV all over the country and they were handling all this blood and it's splashing around the laboratory and (INAUDIBLE) wearing gloves. Now you see New York ME's office DNA unit is really a first class enterprise. It's light years above what it was before.

ARVIZU: Some forensic laboratories are far ahead of others. There are a few in the country that are making very legitimate and very concentrated efforts to improve their quality assurance programs.

UNIDENTIFIED MALE: Dramatic improvement like that seen at the New York lab has also occurred in other labs in the state. The changes began when New York became the first state to regulate crime labs and require they be accredited.

SHALLER: The state has allocated funds to make sure that the laboratories can maintain their accreditation and this is huge because the laboratories now can say, look, I've got this problem. I need money to correct it.

UNIDENTIFIED MALE: Only three states require that crime labs be accredited, so it is mostly a voluntary system. Federal prosecutor Kenneth Melson (ph) is on the board of the agency that accredits labs, now as ASCLADLAB (ph). He says that less than 10 percent of labs earn accreditation on their first inspection. And these are labs that have been processing evidence for years, evidence presented to juries as unassailable fact.

UNIDENTIFIED MALE: So what that means is that many of them have certain failings or deficiencies where we have to go back in and remediate. I have seen tremendous improvement in a crime laboratory from the first time it applies to the time that it finally gets accreditation.

UNIDENTIFIED MALE: Melson estimates that about half the crime labs in the U.S. have earned accreditation. But critics point out that being accredited by your peers is not the same as being regulated by outsiders and it is not a panacea.

ARVIZU: Accredited laboratories have had their operations suspended due to fraudulent practices. Accreditation is no gold star seal of approval. Accreditation is no guarantee of quality.

THOMPSON: We've had some problems with accredited laboratories but the fact that they're accredited helps resolve the issues much more quickly, because part of the quality assurance management program they have is a way to respond to problems.

UNIDENTIFIED MALE: Accreditation is a first step and as scandals unfold across the country, it's clear change is needed.

DAVID FAIGMAN, UNIV. OF CALIFORNIA HASTINGS COLLEGE OF THE LAW: That's really the bottom line here. We're all on the same side. Nobody is in favor of criminals. Nobody is in favor of simply throwing out evidence and not doing anything about it. What we're about here is saying to government agencies, saying to state agencies is, you can do better.

UNIDENTIFIED MALE: Better, so that juries can trust what they hear in a courtroom, so that innocent people are not imprisoned and so that real justice is served.

UNIDENTIFIED MALE: We get locked up with only - it wasn't fair to me and it wasn't fair to his little girl, because the guy that molested her has been out there all these years. She thought she was safe. She wasn't because he was still out there.

TO ORDER A VIDEO OF THIS TRANSCRIPT, PLEASE CALL 800-CNN-NEWS OR USE OUR SECURE ONLINE ORDER FORM LOCATED AT www.fdch.com