A governance framework for defining ethical boundaries in enterprise AI systems

Organizations deploying AI face a question regulation cannot answer: what ethical boundaries will your enterprise tolerate? In Boundaries of Tolerance: A Foundational Model for Ethical AI Strategy and Governance, Jeffrey Saviano—MIT Sloan faculty, Harvard AI ethics researcher, and former EY emerging technology strategy & governance leader—introduces a governance model developed through original research and informed by advisory work with corporate boards and senior executives to help leaders define ethical risk thresholds.

Boundaries of Tolerance presents the author's proprietary framework built on a six-level Ethical Maturity Stack, two Performance Measures, and Applied Governance Practices. Together, they help organizations assess their ethical maturity and chart a path forward. Enriched by case studies of companies such as JPMorgan Chase, and applied practices from Intel to the American Bar Association, the book provides actionable tools for embedding values into AI systems and aligning innovation with integrity.

Readers will also find:

• Governance strategies that empower boards and senior leadership teams to establish clear accountability structures for AI-driven decisions
• A complementary approach to AI strategy built on core elements that guide how leaders define purpose, determine where AI belongs, and maintain performance as conditions change
• Case studies illustrating both strong ethical AI governance and the consequences when AI systems cause harm
• Practical tools for aligning AI innovation with organizational values without imposing rigid compliance-only frameworks
• A leadership blueprint for building stakeholder trust and resilience before litigation or regulatory action forces reactive governance measures

Written for corporate directors, C-suite executives, business professionals supporting those teams, civil society leaders, and policymakers navigating AI adoption, Boundaries of Tolerance provides a structured, research-backed governance model that helps leaders proactively define, measure, and govern the ethical boundaries of their AI systems, including the fiduciary oversight responsibilities that accompany AI deployment.