Even within the tech sector, there are cynics who will claim artificial intelligence is not as effective as thought to be.
But each day brings a new innovation where AI has been able to achieve real breakthrough.
The latest is in predicting that a cyberattack is imminent.
That would come as a big boon for the cybersecurity experts and systems professionals tasked with ensuring the security of the computer network under their charge.
‘DARKMENTION’ Is the Name Given to the New System
Intelligence Advanced Research Projects Activity (IARPA) is a U.S. government-supported initiative and under its aegis, a new program, Cyber-Attack Automated Unconventional Sensor Environment (known by the acronym CAUSE) has been developed.
It is this CAUSE program that has come up with the system that has been given the name “DARKMENTION,” and its primary objective is to apply artificial intelligence to make predictions that a network could be headed for a cyberattack.
The team that worked on developing this program has used a few known and some new concepts in developing algorithms that can keep looking out for any unusual activity within a network and issue a warning on an oncoming cyberattack.
They released a paper [PDF] explaining their findings.
According to the paper, two mapping techniques—Common Vulnerability Enumeration (CVE) and Common Platform Enumeration (CPE)—that identify the vulnerabilities within the software/hardware environment have been used as the starting points to build DARKMENTION.
AI-Based Analytical Tool
What follows when the program DARKMENTION is deployed in a network is after it does the analysis of the different set of programs running within the network, it will dig deep into the data to come up with parallels of unusual activity or anything going on within the system that falls into the pattern resembling a preamble to a cyberattack.
This will trigger a warning to be issued to the user or the systems controller, and that individual can set into motion steps required to stop the cyberattack
DARKMENTION has been deployed and tested by the team that developed the program under different conditions and verified that it works perfectly.
DARKMENTION would be fed to enable it to sound out the warning of an impending attack.
Could Be a Game-Changer
The current scenario, as far as cybersecurity is concerned, is that the hacker does the damage or a malware already makes an entry and then it gets detected, removed and any specific tasks required to prevent similar attacks are carried out. It’s mostly after the event.
To be fair, there are ways some specialists in the field who offer white hat hacking tests can confirm if a network has vulnerabilities.
This is something like the destructive tests done on products.
Not every organization may be willing to spend on such exercises.
It is in this background that a program like DARKMENTION could be a real game-changing development in the industry, and networks owners may be wanting to grab the technology and deploy it in their systems/networks.
If the systems engineers need to be trained on the program and to understand how it works and what kind of warnings can be seen, that may become possible.
You Still Have to Work on the Basics
The tendency among the user-community is to forget the previously implemented precautions, the moment a program like DARKMENTION is deployed.
That might be the wrong thing to do. The best way would be to first do a thorough check on the list of things needed to keep your computers fully secured.
The involvement of all those who use the systems will have to be implemented in too.
There should be no compromise on that.
Once that is done, you may be better equipped to deal with any situation.
The use of this artificial intelligence-based DARKMENTION will come in as an additional help in keeping the hackers away.
There is no denying the fact that the cost of preventing a cyberattack is much less than experiencing one and then taking the remedial action.
One my have to wait a little for this program to be made into a commercially viable product for use by all users.