How AI Increases Cyber Security

This post is also available in: עברית (Hebrew)

Earlier this year, the United States Defense Department has awarded a 20- month contract to a company that develops Artificial Intelligence technology that continuously authenticates mobile users based on their behavior.

Under the terms of a $2.42 million contract, TWOSENSE.AI will be working to deploy deep neural networks for multi factor authentication on mobile devices. The technology works by continuously monitoring and learning user behaviors. The AI learns and memorizes certain behaviors such as keystrokes, fingertip pressure on the device, and gait. If the AI notices something irregular, it will then ask the user to enter their standard login information.

This technology does not replace traditional login authentication, instead it strengthens it.

The company has been working on the technology along with the Defense Information Systems Agency (DISA). DISA has been searching for authentication alternatives to the Common Access Card, an ID card with an embedded chip that allows for login authentication to systems. The user would plug the card into a PC or phone, then enter a PIN to gain access to the system. Once logged in, the user can use the system until the card is removed or a timeout is reached. The user then would need to reauthenticate in order to re-login to the system.

So the new AI technology allows for Continuous Multi Factor Authentication (CMFA). CMFA is a great time saver for the user, since with it there is no need to reauthenticate every time the user’s login has reached a timeout. The user simply plugs in his Common Access Card and uses the system, while AI continuously and non disruptively makes sure that the same user is using the device.

The technology utilizes the sensors that exist on mobile devices in order to work. Accelerometers, gyroscopes, compasses, light level, temperature and air pressure are all sensors used in the authentication process. Wi-Fi and Bluetooth are also used to show the connectivity environment.

TWOSENSE.AI has to deal with a few challenges in order to supply continuous multi factor authentication. One challenge is the amount of bandwidth needed to handle the large amount of data being transmitted, because of this the technology needs access to the cloud network.

Another challenge is to minimize battery consumption. In order to save energy, the technology needs to sync data on Wi-Fi and know when to connect to other networks whenever they are available.

Gcn.com reports that company’s executives believe that one day  the technology will become a standard layer in cybersecurity of mobile devices.