SBIR/STTR Award attributes
Cybercrimes are being committed constantly online against all persons, organizations and nations; becoming one of mankind’s greatest problems; threatening with cyber thefts of information, data, secrets; causing damages estimated in $ trillions; and creating new forms of extortions and crimes (e.g. ransomware). The same dangers could impact our nation’s federal and defense computing systems, which are at risk with grave consequences if their security is compromised. Unfortunately current countermeasures using conventional protection methods have difficulties keeping up with new, sophisticated and mutated forms of adversarial cyberattacks. The Phase-I project develops the Trio System for cybersecurity with advanced artificial-intelligence (AI) technologies that enable self-learning for absorbing relevant data, automating detections of cyber threats, known or mutated or newly devised, and empowering organizations, federal and defense departments to more effectively identify threats. The technologies to be developed include (classical) Machine Learning (ML) and Neural Network (NN) algorithms to provide a modern mechanism for the detection and prevention of cyber intrusions and anomalies in all enterprise and industrial-control-system networks. In addition, the developed self-learning and automation features can support the tasks of relieving cybersecurity personnel from having to sort through large sets of detection results, which presently can lead to “alert fatigue”.
