What if an app on your smartphone could verify that someone is telling you the truth? That seems to be the idea behind an AI lie detector called VerifEye, which has been making the rounds for at least the past year. The app was created by Converus, and the company claims it can detect lies with up to 80 percent accuracy.
The app recently appeared in some Reddit threats on an AI-focused subreddit, and while Converus never explicitly claims AI powers the app, it’s very clear that it uses some type of AI to detect the behavior of your eye, which is what VerifEye focuses on to determine if you’re telling the truth or not.
The AI lie detector looks for involuntary movements and behavior within your eyes to try to determine whether or not you are being truthful. Converus never explicitly says what kind of behavior it is looking for that I could see, at least not directly on the website’s main pages. However, the behavior it focuses on can be detected by simply pointing your smartphone’s camera at your eyes and using the VerifEye app.
The idea that the eyes can tell if you’re lying isn’t a new one. It’s actually been around since the 1970s in some way or another, though one of the most common theories about eye movement was actually disproven in 2012. Still, the eyes and their behavior remain key indicators that you can use (alongside other indicators) to tell if someone is lying. This isn’t the formula that VerifEye follows, though, and a ton of additional research into the science of using ocular data to detect deception can be found on the app’s website.
You can see an infomercial about how VerifEye is supposed to work in the video I’ve embedded above. What’s most interesting to me is how the company advertises this app, as if it’s something we’re all just going to be using in our daily lives. It isn’t just advertised as an AI lie detector for corporations. The company also mentions using it to determine if someone’s online persona is real or even if someone has been unfaithful.
Of course, the claims of 80 percent accuracy aren’t likely to be anything that is admissible in court, especially with the E.U.’s recent AI Act changing up a lot of things. This app might even be considered illegal in some places because of the implications behind it. But this is an ever developing part of the industry right now, so that could very well change in the future.
Still, it isn’t surprising to see a company taking this stance, though I do have some doubts about how useful the app claims it could be. Those doubts and questions haven’t stopped others from jumping at the very hint of an oppressive future where AI is constantly watching our every move. Of course, all of that is only compounded by the fact that AI knows how to deceive humans.