Artificial intelligence and democracy: The road ahead

4.4. Artificial intelligence and democracy: The road ahead

Looking back, we see many areas of the potential impact of artificial intelligence on democracy. While many AI applications and their subsequent effects still lie in the future, we already see AI's impact on several areas connected to democracy. Accordingly, political science needs to examine more systematically how AI is deployed in these contact areas and connect patterns and trends to core concerns of democratic theory and democratic quality.

As we have seen it is important not get sidetracked by grandiose but ultimately imaginary visions of a general artificial intelligence. Instead, we should focus on specific instances of narrow AI, their uses in specific areas of interest, and their effects. The discussion has shown that AI does not necessarily has to contribute to democratic decline. Still, there are many areas of contact, in which we need to examine AI's uses and effects. These are the function of elections as "organized uncertainty" about a temporal settlement of political conflict, the ability of people to decide themselves about who should govern them, equality within societies, and the competition between democracy and other systems of political rule. These areas can serve as focus points where we can look for effects of AI on democracy.

But where does this leave the young enterprising social scientist interested in effects of AI on society and politics but not trained in developing AI itself? There are many research opportunities in this area and far too few people work on this in a serious manner. In other words: opportunities abound.

On a very basic level, there is need for more systematic work on how AI is employed in politics and by governments. There are a few early studies—for example in the area of predictive policing or the legal system—but they are comparatively few and predominantly focused on the US. This provides opportunities for local cases in different context from the US or for comparative work. Much of what we think we know about AI in society is based on journalistic accounts. While those can be highly instructive, there is consistent need for systematic scientific accounts of these developments in different sectors and societies.

There is also a consistent stream of studies examining the biases of specific AI systems in specific contexts. This could be the auditing of AI-driven analysis of texts and images or the implementation of AI in societal processes—such as policing, judicial sentencing, or credit decisions. In a twist on this template, you can also use AI to study historical biases in society by examining historical large scale data sets of texts or images and identify the distribution and shifts in the representation of specific societal groups—for example along gender or racial lines.

Building on this, there is also the need for systematic work on the regulation of AI in different countries and contexts. While these are admittedly early days for the official regulation of AI, there are interesting early activities. One surely is the government induced support for AI-focused research and development. Be it in the US, the EU, or China, governments have identified AI as area of strategic geopolitical and economical competition. Examining competing policy initiatives trying to drive innovation is a promising early step in examining governments' relation with AI and related policy and regulation. Beyond this, earlier comparative studies regarding the regulation of privacy or platform business models can offer templates for subsequent work with a focus on AI.

Finally, the development of AI as a scientific and commercial field is a promising subject of research itself. The highly interdisciplinary nature and shifts in the vanguard of AI development between scientific disciplines, academy and business, and geographic areas between USA, Canada, and China makes this into a very interesting case of the contemporary science, commerce, society nexus. Going further, AI is increasingly used within the social sciences as a method of discovery. The associated shifts in scientific workflow, questions asked, and subsequent approaches to theorizing are a fertile area of future reflection.

In the final account, this shows that there are many promising avenues for future scientific work on the impact of AI on democracy. Here it is important to leave field specific blinders behind. Political science might be slow in addressing these questions but other scientific fields are already hard at work. Many of these questions are already adressed by computer scientists, ethicists, sociologists, legal scholars, and communication scientists. It is important to recognize and build on this work irrespective of its origin. As we have seen, this interdisciplinary work adresses many questions relevant to democracy. This means it is high time for political scientists to add their voice and perspective to the ongoing debate about the impact of AI on democracy.