AI has the capacity to maximise the social, economic, and environmental benefits of big data sources for decision-making. Researchers, governments, companies, non-governmental organisations, and citizen groups are actively experimenting, innovating, and adapting AI tools to tackle policy relevant problems. Research and development are needed to properly design sound theoretical methods and applied tools for generating policy relevant information, implementing policy objectives, designing more effective policies, or measuring their impact while mitigating potential risks. This panel discussion aims at providing some elements to help answering the following questions:
- What are the opportunities created through AI-driven decision-making?
What are the challenges to truly realise the potential of this emerging field?
- How to find the right balance between human and machine (algorithmic) decision-making?
- How to translate the central features of ethical AI, e.g., that it is safe, transparent, unbiased, explainable and accountable, into actions? How to preserve privacy while leveraging the predictive power of algorithmic decision-making?
- What policy fields are best suitable for AI action? How to engage citizens in the process and what are the conditions to ensure trust and acceptance of AI in a policy context?
- Can facts and figures produced by AI be accountable? How to achieve reproducibility by design?
- How to rethink the public sector with AI? What are the role and influence of the main actors in AI from the private sector in the design of public policy?
- What are the policies regarding AI: policy for AI versus AI for policy?