QinetiQ Unveils PrimAITE Software in a Bid to Fortify UK's Cyber Defence

The Defence Science and Technology Laboratory (Dstl) and QinetiQ Training & Simulation have released an open-source digital training ground for "Artificial Intelligence agents" named PrimAITE.

This move is part of Dstl's Autonomous Resilient Cyber Defence (ARCD) program and is intended to address the growing threat of cyber-attacks faced by the UK's Armed Forces.

PrimAITE, or Primary-level AI Training Environment, is now available for public access, aiming to encourage collaboration among governmental agencies and the wider developer community. Developed through the Serapis SSE framework by a team led by QinetiQ, PrimAITE is reportedly in demand across various ARCD projects.

The software serves as a tool for training and evaluating AI and machine learning in a cyber-defensive role. Notably, PrimAITE's customizable environment can model platforms and system contexts, incorporating details such as connections, IP addresses, ports, traffic loading, operating systems, and services. Operating at machine-speed allows for swift training cycles.

PrimAITE aims to contribute to the goal of achieving autonomous cyber defence, enabling military platforms to automatically identify and defend against cyber threats. This is expected to reduce the need for human intervention and enhance response times for military systems.

Craig Murphy, Dstl’s technical lead for ARCD, mentioned that PrimAITE is currently in use by more than 20 projects, including other programs within Dstl and with international partners. Efforts are ongoing to develop the software further.

Version 2.0.0 of PrimAITE is available on GitHub, with plans for version 3.0.0 underway. The QinetiQ Training & Simulation-designed capability, owned by Dstl and the Ministry of Defence, has received support from specialists at Methods Analytics, Envitia, and QinetiQ.