* New software believed to spot liars based solely on their mouse movements
* The AI is said to work with 95% accuracy
* Researchers hope to use the AI to help detect identity theft
With more than 15 million cases reported every year, security technicians are constantly working on creating better, more effective ways of detecting online identity theft. Now, a new program promises to spot liars simply based on how they move their mouse.
A group of Italian researchers led by Giuseppe Sartori asked 20 volunteers to assume a fake identity. From there, the volunteers were asked a series of yes or no questions about their made-up backgrounds.
The researchers then asked the same set of questions to another set of volunteers, only this time the volunteers were telling the truth. The queries ranged from simple questions like where a person was born to more complex questions regarding their alleged zodiac sign.
Sartori and his team used an AI software to match the honest answers to the dishonest ones, finding underlying themes based on the subjects’ mouse movements. In fact, they were able to check the way a person moves their mouse to infer when they were lying and when they were telling the truth with a staggering 95% accuracy rate.
According to the researchers:
“While truth-tellers easily verify questions involving the zodiac, liars do not have the zodiac immediately available, and they have to compute it for a correct verification. This lack of automaticity is reflected in the mouse movements used to record the responses as well as in the number of errors.”
As the biggest problem with detecting online identity theft today is effectively being able to match a person to a specific account, the researchers believe this new software could be the first step toward improving online verification methods.