Despite the use of finger prints, facial recognition, voice recognition, the use of social media to deliver legal documents and other tech innovations in legal settings, the question is to what degree society is prepared to trust this latest, even more intrusive innovation to assist in judging guilt or innocence. JL
Heather Mongilio reports in PBS:
It’s not a question of whether neuroscientists can detect if someone is lying using a fMRI but whether the court system is willing to let the technology in. Experiments suggest that the fMRI machines can detect lies. “For any algorithm-based lie detection that could lead to conviction or acquittal, […] it should be an aid to conviction or acquittal, not the decider. It should not take away the prerogative of judge and jury.”
Inside Gary Smith’s brain was his chance for freedom.
Facing his second trial after appealing his murder conviction, Smith, an Army veteran, attempted to persuade a jury of his innocence by allowing a California company to scan his brain to show he was telling the truth when he said he didn’t kill his roommate.
“For me, it was so important that I do this. It was a very heavy experience,” Smith said.
Lie detection using a functional MRI machine, which measures and creates an image of brain activity, is a topic of controversy among legal and neuroscience experts and has yet to land on the courtroom floor.
Joel Huizenga, CEO at Truthful Brain Corporation, the company involved in scanning Smith’s brain, is hoping to change that with a case in March. For Huizenga, it’s not a question of whether neuroscientists can detect if someone is lying using a fMRI but whether the court system is willing to let the technology in.
Huizenga said that Smith’s brain scan showed that he was telling the truth when he said he did not kill his roommate. He cited multiple studies listed on Truthful Brain’s website, all suggesting that an fMRI can detect whether someone is lying or not.
“A big math problem. That’s all it is,” he said.
And he’s right, to an extent, when he said there’s research backing this up, but it has all been conducted in a laboratory. Dr. Daniel Langleben, out of the University of Pennsylvania, and Dr. Jonathan Hakun, of the Pennsylvania State University, are two that have run experiments using fMRIs for lie detection.
The experiments suggest that the fMRI machines can detect lies, but this lie detection technique has not been properly tested outside of the lab, Langleben said.
In running experiments, looking at fMRI lie detection, Langleben and Hakun looked to see what brain regions are active when people make a choice, Hakun said. The amount of people tested in the different experiments varied, from a single subject to 28 people. Others not conducted by Langleben or Hakun also have smaller sample sizes, ranging from 10 to around 30.
Experiment setups often required the test subject to purposefully lie on yes or no questions, whether it was concealing a playing card or about a number between one and three, and seeing if the machine would indicate if the person was lying, according to a paper Langleben co-wrote in 2012.
Areas in the occipital lobes, which are toward the back of the brain, the parietal lobes, which are toward the middle of the brain, and the prefrontal cortex, at the front, will light up when someone tells a truth or lie. However, they show more activity for a lie, Hakun said.
But the experiments use lies that are relatively simple, not something complicated like claiming innocence in a murder trial, Hakun said.
“There are a lot of ways people lie,” he said.
Langleben compared the use of an fMRI machine for lie detection to self-driving cars. He wouldn’t want to be in a self-driving car without a kill switch, just as he wouldn’t want the fMRI to be the decider of guilt or innocence.
“For any kind of algorithm-based lie detection that could lead to conviction, for example, or acquittal, […] it should be an aid to conviction or acquittal, not the decider,” he said. “It should not take away the prerogative of judge and jury.”
In Hakun’s opinion, there are more questions for fMRI lie detection studies to answer.
“There’s too much to be worked on still to move this into a courtroom,” he said.
The judge in Smith’s case did not admit his brain scan, which indicated he was telling the truth, as evidence, a decision that did not surprise Smith because it was such a new technique. Smith’s case was eventually appealed again, and he took an Alford plea, where he maintained his innocence while admitting the prosecution had enough for a guilty verdict.
Smith said he still believes in the credibility of the test and is willing to be an advocate for the fMRI lie detection or a test subject. He currently works for his attorney as a law clerk while also attending college.
“I took [the fMRI lie detection test] in the hope of being able to say, not just to the prosecutor, but to my buddy’s family, ‘hey, I didn’t do this,’” Smith said.
0 comments:
Post a Comment