In order to convict someone of a crime, the evidence has to show beyond a reasonable doubt that the suspect is guilty of all the elements of the offense. One basis to help obtain sufficient evidence to prove the case beyond a reasonable doubt is where the suspect admits to the crime. When the police first question a suspect, he or she rarely admits to being guilty. However, it is not easy to tell if the person is guilty but lying, or the person is innocent and telling the truth. The police may claim they have a “hunch” but a hunch will not stand up in court.

Police in Spain have recently reported success with an algorithm that they are using to identify when people are lying to the police. While this formula may help statistically identify more common instances of false reports, it will not likely be admissible for any criminal cases in the U.S. anytime soon.

Do Lie Detectors Work?

Some people think a lie detector can help identify when someone is lying or telling the truth. However, these devices are not a reliable test of truth-telling and are generally not admissible as evidence in a criminal trial.

Lie detectors, or polygraphs, cannot measure lies. Instead, they test for physiological reactions to the testing situation, including increased heart rate, increased blood pressure, breathing rate, or perspiration. Some people may show increased rates when they are lying. However, people may show increased rates when they are nervous, stressed, or have anxiety. Certain medical conditions or medications could also affect a lie detector test.  

Police in Spain Testing False Robbery Reports

In Southern Spain, police departments were concerned that they were spending too much time responding to robbery reports that were not real. According to the police, people were falsely reporting robberies to collect insurance money or because individuals did not want to tell spouses or family members that they'd lost or gambled away money.  

A police inspector and mathematician, Miguel Camacho Collados, along with other scientists, developed an algorithm to try and identify false robbery reports based on the wording of the reporting statements. The VeriPol system was tested by analyzing closed cases that had ended in either a conviction or the reporter confessing to fabricating the robbery. When compared to human evaluation, the VeriPol system outperformed the human experts.

One of the patterns of false reports included using shorter sentences, lack of information about the actual robbery, and certain types of attacks (like the attacker had a helmet or attacked from behind).

Using Computers to Test Criminal Suspects

Using a computer to test for false robbery reports may help police identify individuals who may statistically be less likely to be reporting the truth. However, this could result in the police failing to take real reports as seriously, just because the victim aligned with the false report profile.

Similarly, using similar artificial intelligence (AI) to test for lying in a criminal suspect would not likely meet the threshold of “beyond a reasonable doubt.” Each criminal case is different and each individual facing the prospect of jail time, loss of freedom, and loss of their rights should be given the full deference of the criminal standard of proof provided by the Constitution.

Charlottesville Criminal Defense Lawyer

One of the tricks used by police and prosecutors is making people think they have no choice but to plead guilty. The police may make it seem like the suspect is lying, like they have a strong case, and if the person admits to their guilt, they will get a lesser sentence. However, before ever saying anything that could be used against you, talk to a lawyer. It is your constitutional right.

Don't let the police or prosecutor decide whether you should plead guilty or not. You may have a stronger case than you realize. Your Charlottesville criminal defense attorney will investigate your case, identify all your defenses, and challenge the criminal case against you. Contact me today for a free consultation.