INTEL is working on a way of protecting laptops from malicious activity by taking into account how people use their computers in different ways.

Standard protection software that triggers an alert if the amounts of traffic being sent out move above a certain threshold are flawed, Intel reckons, because not everyone behaves in the same way with regards to the amounts of data they send out or receive.

Intel’s has been working on a project called Proteus, which is based on a set of algorithms that are capable of making security judgments based on a person’s individual web browsing habits. Basically. it keeps tabs on what you're up to and gets spooked if you do something outrageous

Nina Taft, a researcher at Intel Research Berkeley, told Technology Review that the first algorithm uses standard statistical and machine-learning techniques to observe a user's Web use and create individualised traffic thresholds. The second algorithm asses how a user’s Internet behaviour changes throughout the day. Then, a third set of algorithms, for detecting whether or not a computer has been hijacked and turned into a botnet, examines the communication between a person’s laptop and other machines on the Internet.

Taft reckons that by creating protection software which adapts itself to various user profiles accurately, hackers will find it more difficult to carry out their malicious behaviour undetected. The idea of using behavioural data to make security software more accurate isn’t a new one, but has so far mainly been limited to routers that monitor network activity. Intel’s software would be aimed at company laptops and smartphones.

Intel also hasn’t ruled out the possibility of hardwiring Proteus to a computer’s circuitry. According to Taft, "Intel is interested in getting as much [security] into hardware as possible". µ

L'Inq
Technology Review