Trust is not only a regulatory requirement but also a prerequisite for widespread deployment and value creation. Europe should continue promoting trustworthy AI that combines technical robustness with legal certainty, transparency, and human oversight. This requires combining technical safeguards, governance mechanisms and appropriate regulatory frameworks throughout the AI lifecycle. Priority areas include:
- Ensuring safe and reliable deployment at scale, strengthening the robustness, resilience, security and reliability of AI systems in real-world environments and high-impact applications.
- Technical and governance safeguards against risks and misuse, combining technical measures, methodologies and legal frameworks to identify, mitigate and monitor AI-related risks and prevent harmful or unintended uses.
- Responsible and human-centric AI by design, embedding values such as transparency, fairness, accountability and human oversight throughout the AI lifecycle, from data and model development to deployment and continuous monitoring.