What does someone look like when they lie? Their face might turn red, they may stare at the ground, or maybe they’re good at it. You might not even notice unless you were looking at a polygraph test.
But how do you tell if someone is lying online? Tons of sensitive information gets passed over servers everyday, and knowing if a user is trying to be deceitful would be extremely valuable to many businesses. Thats where technology like Neuro-ID becomes beneficial to companies that handle that information.
Neuro-ID, a start-up company founded by Joe Valacich, a professor of management integration systems in University of Arizona’s Eller College of Management, and one of his former grad students, does exactly that. It uses information gleaned for your mouse clicks, speed, acceleration, time stamps and coordinates, in order to measure physiological responses to indecision and noticing malicious intent.
“Your cognitive responses, anger, confusion, they manifest as measurable changes in the nervous system,” Valacich said. “The way you move changes.” The software can use the changes in your mouse movements to flag people who may be attempting to be fraudulent on certain forms, for example insurance claims or applying for a passport.
“A lot of security is geared towards identifying the intent of what someone is trying to do,” said Michael Byrd, senior vice president of Neuro-ID. Even common spam filters are trying to figure out the intent of an email; they simply flag keywords in text rather than mouse movements.
There are public concerns about systems like these being invasive. Cooper Quintin, a staff technologist at Electronic Frontier Foundation, a non-profit that advocates for online privacy rights, says that the possibility of systems like these flagging false positives could be detrimental.
The technology that Neuro-ID has developed, has a six to seven percent of people who are flagged as false positives, according to Valacich. The risk of being wrongly thought of as lying on forms as important as health insurance or other official documents is excessive for Quintin.
“The consequences of flagging an email as spam are far less than the consequences of a false positive where you have the assumption that they’re committing fraud,” Quintin said. Although being flagged by Neuro-ID would not automatically assume you are guilty of some crime or prevent you from receiving insurance, the thought that your intent is being judged by your mouse movements has negative impacts, according to Quintin
Systems like this that track your mouse movements and tries to measure your intent have been in use for a while in different capacities. Security that tries to recognize click fraud, a common way of advertising companies generating more income by manipulating the amount of views their advertisements receive, already tracks user’s mouse movements.
“It’s simply finer grain web analytics,” Valacich said. “There is no definitive lie detector, there is no crystal ball. What we detect is when people have indecision or some sort of reaction to a particular question that was different than answers to benign questions.”
An example of this would be filling out a medical insurance questionnaire, confirming your name and address would be a no-brainer; you wouldn’t hesitate. If the user hesitates every time a question about tobacco use comes up and it’s noticed by the program in the way the user’s mouse moves, there might be fraudulent behavior.
Although the majority of users are being honest, Valacich and his team are trying to find the extremely small percentage of people who are being dishonest. There’s no manual on where to start looking
“If people have a change in their motor movements due to changes in their cognitive state, that might be a good place to look,” said Valacich. “It’s an additional way to provide insight so you can prioritize your investigative resources.”
While users may not know it, this is the kind of data that is being collected when you agree to the terms of service on almost every website you use.
“People are not aware that while you’re reading things in your browser, websites can actually monitor your behavior,” said Quintin. “People think of it as a one-way communication medium, but it’s a two-way medium.”
Tanner Clinch is a reporter for Arizona Sonora News, a service from the School of Journalism with the University of Arizona. Contact him at [email protected]
Click here for a word version of this story and high resolution photos