Ethical Voice in the Age of Artificial Intelligence: When AI Participates in Decision-Making, Who Has the Right to Disagree?

This research paper examines the protection of ethical voice in workplaces where artificial intelligence (AI) supports or influences decisions affecting employees. It focuses on a critical human-centred question: when AI participates in recruitment, performance evaluation, task allocation, customer service, or managerial decision-making, do employees have the right and the psychological safety to understand, question, challenge, and seek correction of AI-supported decisions? The paper analyses the relationship between psychological safety, AI transparency, meaningful human oversight, and employees’ willingness to raise concerns or challenge potentially unfair or harmful automated decisions. It draws on academic literature, international governance frameworks, regulatory guidance, and documented real-world cases, including AI bias in recruitment, algorithmic grading, automated welfare fraud detection, and workplace monitoring. The study identifies major governance risks, including lack of transparency, automation bias, weak accountability, absence of safe grievance mechanisms, and insufficient human review. It proposes the Ethical Voice Model for AI Governance, based on four pillars: transparency, fairness, human oversight, and accountability and appeal. The paper concludes that protecting ethical voice is essential for responsible AI governance, workplace fairness, human dignity, and sustainable institutional trust. It provides practical recommendations for organisations, policymakers, compliance professionals, and AI governance practitioners.