Hackers could now know what people type on Zoom video call by evaluating the shoulder movement of users

In the time of great distress when COVID-19 pandemic became the cause of lockdown around the world due to which all the type of professional work is done online using certain platforms such as, Zoom and Google Chats.

From the research done by the University of Texas at San Antonio and the University of Oklahoma, it was demonstrated that people's passwords and other things that they type during a video conference meeting is not safe as Hackers could be able to guess their passwords and much more things by analyzing the body position and language of the person while on a video conferencing call.

According to their research, hacker could be able to guess keystrokes whilst on a video conference or meeting call by analyzing the movement of the shoulders while typying, these gestures allows hackers to identify and trace the specific entries.

Researchers had found out that almost every famous video calling platform which is used worldwide has been affected by this and the research team has also found out that the keys that are pressed by the user during using these platforms can be detected by analyzing the upper arm movement of the users with almost 93% accuracy.

According to the University of Texas assistant professor of computer science Murtuza Jadliwala which was the person who leads this research stating its concerns about the possibilities which could happen if hackers could get access to the information which is classified. He states that this problem is a very big concern in a broader perspective, which was been overlooked for a long time.

On the occasion, he also said that the reason for him to initiate this research was not COVID-19 and this research took a year but he came to the conclusion that due to COVID-19 the risk of such attack is higher than the usual as now mostly the work is done online.

His research team evaluates that during typing there are subtle pixel shifts around the user's shoulder to identify the four main directions in which the user was moving. Four main directions are North, South, East and West.

This movement allowed researchers to know in which direction the user is moving to type on the keyboard.

From all the information using previous research, researchers developed a program that made word profiles using an English dictionary from the sequence of movements.

It was an alarming situation for developers that they have not used any sophisticated machine learning or AI technology to develop such a program, so this could be very easy for hackers to use this information against users.

During their research, certain issues were encountered by them some of which are listed below:
  • In the lab, it was noted that the average accuracy was around 75%
  • System struggles with long sleeves rather than short sleeves
  • System faces problems when user have long hair covering their shoulders
  • It's difficult for the system to act for slow typers rather than fast typers.
Jadliwala warns these video calling platforms about this and urges them to ensure their users protection.


No comments:

Post a Comment