Accountability & Governance

About This Module

When an AI system causes harm, who is answerable? A model cannot be held responsible, so the responsibility has to sit with people and organisations: the developers who build it, the companies that deploy it and the regulators who oversee it. Accountability means that someone is clearly answerable for an AI system’s outcomes and can be asked to explain them and put things right. AI governance is the set of policies, roles, processes and oversight that make that accountability real.

The lessons begin with why accountability matters and what it looks like for leaders. They then cover practical structures: the NIST AI Risk Management Framework, which organises risk work into four functions (Govern, Map, Measure and Manage), and what building a responsible AI programme involves in an organisation, including audit and compliance.

The final lesson looks at independent auditing. Outside researchers have uncovered many AI harms, yet they often lack access or legal protection. A proposal for third-party auditor access, and the discussion that follows it, shows why accountability depends on law and institutions as well as good intentions. Watch the lessons in order, then continue on to AI Safety & Alignment.