The Hebrew University of Jerusalem is proud to announce its participation in RobustifAI, a groundbreaking Horizon Europe research consortium dedicated to strengthening the reliability and robustness of Generative Artificial Intelligence (GenAI) technologies. The project officially commenced on June 1, 2025, with a total budget of €9.3 million and a projected duration of 36 months.

Led by a diverse international coalition of academic institutions, research organizations, SMEs, and industry leaders, RobustifAI aims to develop a rigorous, human-centric methodology for designing and deploying trustworthy GenAI systems. This effort is particularly focused on the integration of neural and symbolic methods within human cyber-physical systems (HCPS)—complex systems that merge computational processes with real-world human and environmental interaction.

Pioneering Research in Human-Centric AI
Professor Guy Katz of the Hebrew University’s School of Computer Science and Engineering will lead the university's efforts in the consortium. His expertise in verification and formal methods is central to RobustifAI’s goal of addressing the technical, operational, and user-centric vulnerabilities in foundation models.

“Generative AI is a powerful engine for innovation, but to realize its full societal and economic benefits, we must ensure it operates safely, reliably, and ethically—especially when human lives and physical systems are involved,” said Prof. Katz. “RobustifAI is an essential step in that direction.”

Project Launch and Strategic Objectives
The consortium’s official kick-off meeting was held on June 5, 2025, at the AIT Austrian Institute of Technology in Vienna, Austria. Partners discussed the project roadmap and cross-collaboration among the various work packages and research activities.

By targeting GenAI’s use in HCPS—such as autonomous vehicles, healthcare systems, and smart manufacturing—RobustifAI will address the most demanding scenarios where safety and robustness are mission-critical. The methodologies developed will extend to a broad range of sectors, ensuring cross-domain applicability and positive societal impact.

A Strong Global Partnership

The RobustifAI consortium brings together 18 partners from 12 countries, including:

• Seven universities: University of Liverpool, Carl von Ossietzky Universität Oldenburg, Chalmers University of Technology, Hebrew University of Jerusalem, Masaryk University, TU Wien, and Indian Institute of Technology Kanpur.
• Two research and technology organizations (RTOs): AIT Austrian Institute of Technology and German Aerospace Center (DLR).
• Five SMEs: CSX-AI, L-up, PPM Robotics, CertX, and LOXO.
• Four industry leaders: Collins Aerospace, Siemens, Thales, and Thales SIX GTS France.

Together, this consortium will spearhead efforts to embed robustness, explainability, and human-centered design at the core of next-generation AI technologies.