Ethics and Governance of Autonomous AI Agents Training Course

Core Ethical Principles in Autonomous Systems

• Defining the concept of autonomy in AI agents
• Applying key ethical theories to machine behaviour
• Incorporating stakeholder views and value-sensitive design principles

Societal Implications and High-Risk Applications

• The role of autonomous agents in public safety, healthcare, and defence
• Navigating human-AI collaboration and establishing trust boundaries
• Addressing scenarios of unintended outcomes and amplified risks

The Legal and Regulatory Environment

• Survey of AI laws and policy developments (including the EU AI Act, NIST, and OECD guidelines)
• Issues of accountability, liability, and the legal status of AI agents
• Reviewing global governance efforts and identifying regulatory gaps

Clarity in Decision-Making and Transparency

• Overcoming the challenges of opaque, black-box decision-making
• Creating agents that are explainable and subject to audit
• Utilising transparency tools and frameworks such as model cards and datasheets

Alignment, Control, and Ethical Duty

• Implementing AI alignment strategies to guide agent behaviour
• Comparing human-in-the-loop and human-on-the-loop control methods
• Defining shared responsibilities among designers, users, and institutions

Evaluating and Managing Ethical Risks

• Mapping risks and analysing potential failures in agent design
• Integrating safeguards and kill-switch mechanisms
• Conducting audits for bias, discrimination, and fairness

Structuring Governance and Institutional Supervision

• Adopting principles of responsible AI governance
• Implementing multistakeholder oversight models and auditing processes
• Building compliance frameworks specific to autonomous agents

Recap and Future Directions