Responsibility with AI

This book is intended for board members who want to use artificial intelligence strategically without being IT specialists themselves. Its starting point is the insight that AI is now part of value creation, while responsibilities, transparency, and liability often remain unclear. The book provides a compact orientation framework for the oversight, steering, and control of AI systems.In the first part, readers receive a board-ready handout: the key opportunities offered by AI, the most important risks - hallucinations, bias, costs, data protection, cyber risks, and organizational risks - as well as three leadership questions: What is already happening with AI within the company? Who bears responsibility when a system makes a decision or an error? How are transparency and traceability ensured? Building on this, the book explains only as much technology as is necessary - how language models, GenAI, and AI agents work - and derives the implications for liability, controllability, and auditability.The focus is on governance: roles within the board of directors and its committees, the AI governance board, AI register, policies, reporting, and integration into existing governance and risk structures. Legal foundations - Articles 716a/717 of the Swiss Code of Obligations, the Swiss Data Protection Act, and the EU AI Act - and their impact on D&O insurance are explained in practical terms. Concrete tools such as a sample AI register, control questions for board meetings, a board playbook with quick wins and 90-day plans, and case studies from Swiss companies support implementation.An outlook to 2030 outlines AI maturity levels, possible audit obligations, AI assurance reports, and the prospect of a Swiss AI governance standard. The book is designed as a leadership tool and is complemented by the companion volume "AI Agents - How Machines Think, Act, and Are Led", which provides the technical background.