Human agency and oversight are essential to trustworthy AI, ensuring it aligns with ethical norms and safeguards fundamental rights. Human agency preserves the autonomy of individuals using or affected by AI, fostering informed and unmanipulated interactions, while also securing the right to rectify AI-driven decisions. In contrast, human oversight entails a supervisory role over AI, involving monitoring and guiding its learning and actions to prevent harm to health, security, safety or rights, especially in high-risk AI scenarios as emphasized in the EU AI Act. These concepts, while closely related, define different levels of human interaction with AI to maintain control and ethical integrity.
