November 09, 2011
A strong password contains capital and lowercase letters, numbers and some special characters. Done properly, the result is a password that grants access to computer systems to the proper user. The only problem is the password is hard to remember, and it’s not supposed to be written on yellow sticky notes that can sometimes be found on the bottom of keyboards. And don’t get comfortable with this long password; it has to be changed every 90 days or so.
With DARPA’s new Active Authentication program, program manager Richard Guidorizzi would like to change that. Instead of current authentication systems that force humans to adapt to computers, Guidorizzi wants to make computers adapt to the humans that built them in the first place. He wants researchers who will work in the Active Authentication program to investigate innovative software approaches that determine a computer user’s identity through activities a user normally performs.
This changes how things are currently done by removing the secret a human holds, a cumbersome and hard to remember password, and focuses on making the user the actual password. Guidorizzi puts it a different way.
“My house key will get you into my house, but the dog in my living room knows you’re not me. No amount of holding up my key and saying you’re me is going to convince my dog you’re who you say you are. My dog knows you don’t look like me, smell like me or act like me. What we want out of this program is to find those things that are unique to you, and not some single aspect of computer security that an adversary can use to compromise your system,” said Guidorizzi.
While these identifying aspects of a person are what we hope to use to grant levels of access to computer systems as appropriate, Active Authentication seeks to make you the key to your access, not to track aspects of who you are. Guidorizzi expects researchers to take special care to ensure this program doesn’t violate privacy laws or allow information about a user’s identity to be misused by others. He doesn’t want to capture user aspects in a database; he only wants to use this information as the key to user access of their computer systems.
Examples of existing research include work with fingerprints, although deployment of sensors makes this more challenging so this program focuses more on software-based solutions. Mouse tracking has received attention as a tool that can validate a person’s identify while sitting at a computer, suggesting this as a possible candidate for further research. In addition, forensic authorship is a field where people are able to identify an author by language usage.
Guidorizzi hosts Active Authentication’s proposers’ day Nov. 18. Those interested in attending can find additional information here.
Please direct all media queries to Outreach@darpa.mil