Thales selected for first European project to develop sovereign AI for embedded cyberdefense
Thales has been selected for the Artificial Intelligence Deployable Agent (AIDA) project funded by the European Commission through the European Defence Fund (EDF).
www.thalesgroup.com
- A total of 28 European industry partners, start-ups and research centres have joined forces on this project to develop a sovereign AI-enabled cybersecurity agent to protect aircraft systems from cyberattacks.
- The goal of this three-and-a-half-year European project is to design an AI with an autonomous or semi-autonomous response capability to provide cybersecurity protection for aircraft systems such as onboard computers and electronic warfare systems on combat aircraft, which are vulnerable to increasingly sophisticated cyberattacks in today’s high-intensity conflicts.
- AIDA is the first European structural framework project in support of the NATO concept of Autonomous Intelligent Cyberdefence Agent (AICA).
Thales is technical coordinator for the AIDA project funded by the European Commission, with CR14 in Estonia in charge of overall project coordination.
This EDF project is a response to three major challenges faced by the armed forces today: attack surfaces are growing due to battlespace digitisation; the cyberattack detection-response chain needs to be automated due to the ever-greater use of autonomous systems such as drones and robots; and AI is being used ever more widely both to launch and respond to cyberattacks.
Responding to the 2023 European Defence Fund call for projects for the development of deployable autonomous AI agents,1 Thales submitted an innovative proposal based on the training of intelligent cyberdefence agents capable of identifying, protecting, detecting and responding to cyberthreats in real time in the five military operating domains:2 land, air, sea, space and cyberspace.
Thales will also lead the project to develop a prototype aircraft using frugal AI agents to protect electronic warfare equipment installed on combat aircraft. This prototype will be tested, using Thales’s Cybels Analytics solution in particular, in scenarios including cyber-electromagnetic threats and advanced adversarial AI attacks.
AI is being used increasingly in the theatre of operations to increase the detection performance of air defence radars, for example, and to help plan tactical missions and assign tasks to swarms of drones and robotic systems. This type of AI must be reliable, robust and cybersafe to prevent it being exploited by hostile forces in any environment (land, sea, air, space and cyberspace). To counter this type of threat, Thales’s Friendly Hacker Unit will conduct a battery of adversarial AI attacks and define appropriate countermeasures to ensure that these cyberdefence AI agents can never become targets themselves.
www.thalesgroup.com
