Pentagon signed the contract for implementation of machine learning in the system of admission to state secret

At the end of May, 2019 the Perspecta Enterprise Solutions company signed with the Pentagon the contract for the amount of $75 million which is directed to upgrade of a security system of department, in particular, on implementation of machine learning.

The security service of the Ministry of Defence and Management of the systems of defense communication of the USA asked Perspecta company to improve information technologies of national help desks to accelerate internal investigations and pronouncement of solutions on the system of admission to state secrets.

The Pentagon signed the contract for implementation of machine learning in the system of admission to a state secret

Perspecta should develop a prototype of the program for continuous check of security and identification of internal threats. Including the prototype will use machine learning and technologies of processing of texts in natural languages to make the file of workers which can be allowed to a state secret. Earlier complete investigative packet prepared manually and was transferred to committee for decision.

The prototype will allow to adapt easily to new technologies and risk categories and also to gradually unroll new verifiabilities. The DevSecOps system will work with the software in Amazon GovCloud cloud for unclassified networks approved by the Ministry of Defence.

The head of the new program of security Terry Carpenter noted that the company considers permanent automatic check as a method of minimization of manual work. Automation reduces costs and provides accuracy of the data. Besides, thanks to implementation of new technologies, it will not be necessary to review admissions any more than each five-ten years as verification process happens in real time and it is based on assessment of advantage and risk. However, while not clearly, such assessment borrows what is the time.^[1]

Notes