By Ben Hartwig, Guest Blogger, Marketing –
Lying is a complex social skill, but it’s not exactly difficult. Motivated people can hide many obvious physical signs of deception, especially during a short interaction. This is a substantial risk when conducting an interview or an investigation.
Governments and businesses have utilized lie detection technology for over a century to aid their decision-making. Today, artificial intelligence (AI) is a major tool in making lie detection more accurate and accessible.
Connection Between AI and Lie Detection
Nearly every industry is working to integrate AI into its operations; lie detection is no exception. AI programs learn by absorbing massive datasets and drawing outcomes based on the results of that data.
For lie detection, this means recording a liar’s minute movements like muscle twitches, vocal fluctuations, and eye vibrations. Machine learning algorithms train on millions of these records to identify the telltale signs of deception. AI can even adjust and learn as the datasets update, so the technology only grows more precise over time.
Techniques and Approaches for Lie Detection
Lie detection has a long history, spanning from the invention of the polygraph in 1921 all the way to the pupil-tracking software employed by solutions like EyeDetect®.
The polygraph test remains the media’s go-to representation of lie detectors. Monitors graph the subject’s stress indicators like heart rate and respiration to identify a lie. However, these tests are crude and ineffective against people with reduced bodily responses to lying. Polygraphs also can’t detect more subtle physical signs like AI programs can.
Tiny changes in our expressions, called micro-expressions, can hint at deception. Often, when we “feel” like someone is lying to us, it’s because we’re subconsciously noticing these changes. However, human eyes miss these signs more often than not.
Previously, researchers tracked micro-expressions by attaching sticker electrodes to the subject’s face. These electrodes detected imperceptible muscle twitches; when certain muscle groups activated, it was a sign of lying.
However, cameras are growing rapidly more sophisticated. Some phone cameras can capture micro-expressions without the help of electrodes. This allows AI programs to analyze these micro-expressions against a historical database and determine who is lying.
The eye-tracking software monitors the movements and patterns of the subject’s eyes during questioning. Using high-powered cameras, researchers can detect when your eyes focus on an object (fixations) or take on a broader field of view (saccades). These eye movements are uncontrollable, unlike the signs observed in a typical polygraph.
Eye tracker algorithms read the eyes, primarily the pupils and fovea, to assess likely cognitive processes such as deception. Common positive signs are a prolonged fixation or rapid switches between fixations and saccades.
Confidence in this technology has risen enough that police departments are using it for investigations. And as with facial tracking, modern smartphone cameras are powerful enough to run the software. These advancements have drastically reduced the cost barrier to applying AI learning to your hiring process.
It can be as simple as downloading apps like VerifEye™ which can accurately detect lies with a quick 10-minute test.
AI is revolutionizing lie detection by providing more accurate, timely, and personalized solutions. Perhaps the most beneficial outcome is that you don’t need a trained professional to understand the readings. AI programs like EyeDetect analyze your subject in real-time, which allows you to focus on other tasks.
With careful implementation, AI-powered lie detection is an invaluable tool in various domains, fostering trust, security, and integrity in our increasingly complex world.
Writer Ben Hartwig is a Web Operations Executive at InfoTracer who takes a wide view from the whole system. He authors guides on entire security posture, both physical and cyber.
Photo by /Hitesh Choudhary