Smartwatches and other wearable devices may make us vulnerable whenever we don them due to their motion sensors.
Tony Beltramelli, an IT University of Copenagen student says in his recent research paper that wearables provide a new “pervasive attack surface” threatening the privacy of its users.
When a person types, whether it is on their smartphone, tablet, building entry system, or ATM, they are moving, hence their movement patterns could be traced back to what their fingers have typed. “Keystroke rhythms are unique enough that they've been used alongside passwords in multifactor authentication,” says Bill Camarda, Naked Security author.
Researchers link motion sensor information with neural networks and machine learning systems to overcome the different device sizes, screen orientation and sensor chips that affect accuracy.
Beltramelli has developed new work with advanced deep learning algorithms that achieve 73 percent accuracy in touchlogging and 59 percent in keylogging—meaning a fraudster can get further using their own methods. “Our results suggest that the complete technological ecosystem of a user can be compromised when a wearable wristband device is worn,” said Beltramelli in his research paper.
Beltramelli advised that the algorithms he used are highly available in open source projects, so anyone can take his work and push forward and they very well may do so.
Camarda pointed out that Smartwatch technology continues to improve with better accuracy as well, giving code crackers more reason to push forward with Beltramelli's research.