Artificial intelligence projects are increasingly expected to deliver multiple outcomes simultaneously: regulatory compliance, financial benefits, innovation, sustainability, and positive societal impact. Yet when these objectives compete, who is actually accountable?

This question sits at the heart of my forthcoming research chapter, “Accountability in Artificial Intelligence (AI) Projects: System Capabilities, Delivery Pressures, and the Limits of Role-Based Governance,” which will appear in the Springer volume Information Technology for Management: AI-Centered Innovation from Theory to Practice: https://link.springer.com/book/9783032270528

The research examines whether accountability in AI projects is truly organized through formal governance structures and role definitions, or whether it is shaped by system capabilities, delivery pressures, and competing organizational priorities.

The graphical abstract summarizes the study’s context, key findings, and practical implications.

Why This Matters

The research suggests that accountability is not a static responsibility assigned once at project initiation. Instead, it evolves dynamically as organizations balance innovation, compliance, business value, and societal expectations.

Understanding these dynamics will be critical for organizations implementing the EU AI Act and other emerging AI governance frameworks.

I look forward to sharing the full chapter when the volume is published and to discussing how practitioners, researchers, and policymakers can strengthen accountability in complex AI initiatives.