This paper considers how artificial intelligence (AI) has profound implications for democracy.

It is not clear yet if the primary impact of AI will prove a liberating function, informing the masses of intelligence that was once the privilege of the few, or whether AI will on the contrary serve as a principal tool of undermining informed democratic participation, reinforcing the power of entrenched technocratic elites. 

AI is here and cannot be ignored by public sector bodies, nor should it be regarded as something transcendental that has to be accepted as a solution to all organizational problems in all circumstances (as tech companies often like to convey). 

The paper examines: 

• the potential of AI to radically disrupt democratic institutions and the erosion of democratic discourse 
• the potential and limits of generative AI chatbots 
• the trajectory and control of technological transformation 
• the concentrated power of big tech and the regulation of big tech and AI
• international policies on the regulation of artificial intelligence 
• artificial intelligence accountability and implementation in the public sector.

Key points

• Selective and appropriate applications of AI will be required to enhance public sector performance. Innovation silos, lack of organizational alignment, and lack of expertise and knowledge will need to be remedied. 
• A coherent vision of the potential of AI in public policy is necessary, with boundary spanning across technical and operational departments, and continuous integration of knowledge into operational capabilities.