Atos and Leading Research and Academia Partners Launch "Forensics of Intelligent Systems", a Joint Research Program to Secure AI against Manipulation

Core Insights - Atos has launched a joint research program titled "Forensics of Intelligent Systems" aimed at securing AI systems against manipulation, in collaboration with various research institutions and the Cyberagentur [1][5][6] Group 1: Project Overview - The research project will develop prototype methods and tools for legally sound detection of manipulations in continuously learning AI systems, including a simulation environment for testing forensic methods [4] - Existing AI models will undergo simulated attacks to identify legally usable traces, and AI algorithms will be adjusted for early detection of manipulations [4] Group 2: Strategic Importance - The initiative enhances state integrity, meets societal expectations for trustworthiness, and increases economic security in a field marked by technical uncertainties [6] - Atos positions itself as a committed partner in AI, focusing on trust, transparency, and European sovereignty through this research project [6] Group 3: Organizational Background - The Cyberagentur, founded in 2020, aims to implement innovative cybersecurity projects with high disruption potential, contributing to Germany's National Security Strategy [7] - Atos Group, with approximately 67,000 employees and annual revenue of around €10 billion, is a leader in digital transformation and cybersecurity [9]