Trustworthy, responsible, ethical AI in manufacturing and supply chains: synthesis and emerging research questions

Manufacturing organizations lack practical guidance on AI risks: While high-level frameworks for responsible AI exist, manufacturing practitioners struggle to understand and implement them in real-world settings, leaving organizations, workers, suppliers, and clients exposed to potential harms.

AI risks span the entire development process: The paper demonstrates through a machine learning lifecycle framework that trustworthiness concerns can arise at every stage of AI implementation in manufacturing—not just in the final deployed system—requiring attention throughout the entire development process.

Multiple stakeholders are affected by manufacturing AI: Responsible AI in manufacturing isn't just an organizational concern; it impacts workers, suppliers, and clients, meaning that ethical AI practices must account for risks across the entire supply chain ecosystem.

There's a gap between theory and practice in responsible AI: Despite various frameworks and definitions for ethical and trustworthy AI being proposed at high levels, the manufacturing sector lacks concrete, actionable approaches to translate these concepts into actual business practices and risk management strategies.

Urgent research is needed to ensure safe AI adoption: The paper calls for targeted research questions to guide the manufacturing community in implementing responsible AI, emphasizing that without this work, the economic and societal benefits of AI in manufacturing cannot be safely realized.