Cybersecurity: When machines and humans collaborate

Machines and humans are good at different things; the former excel at logic and processing, while the latter are more suitable for abstract thinking and comprehensive assessment, but what happens if we combine both?

The Defense Advanced Research Projects Agency (DARPA) wants to find out. That’s why they started an initiative called Computers and Humans Exploring Software Security (CHESS).

The basic idea behind it all is to get the best of both worlds. While the machines never get tired and their logic is flawless, they are still unable to work out some of the more complex problems that humans are better suited for solving. For example, computers are superior at solving math problems, while a complex set of rules such as the syntax of a language is best tackled by humans.

Dustin Fraze, the program’s director, got the idea while watching a DEF CON cybersecurity contest.

The agency will be focusing on achieving the following goals:

– Making systems further protected against cyber-attacks

– Improving situational awareness in cyberspace

– Improving the military’s ability to strike back in cyberspace

There are other DARPA programs as well, some of them focusing on making it easier to build software using “formal methods”. The process takes advantage of mathematical methods to make sure that the code doesn’t do anything it isn’t supposed to.

The CHESS program was launched on April 3 2018, and there was a “Proposers Day” where all the organisations interested in conducting the research were able to participate in a discussion regarding the program’s objectives.