The Forum Network is a space for experts and thought leaders—from around the world and all parts of society— to discuss and develop solutions now and for the future. It aims to foster the fruitful exchange of expertise and perspectives across fields, and opinions expressed do not necessarily represent the views of the OECD.

I didn’t have ChatGPT write this blog for me—but I did have it write it WITH me.

It’s become clear to me that the magic is in the interface, not necessarily the accuracy of the response. It can be a bit of a fabulist, but it will get better, fast. And while issues such as accuracy, privacy and bias must be part of any artificial intelligence (AI) discussion, I was somehow comforted by the fact that the very human act of interpreting the information that the machine delivered hadn’t been crossed out of the equation. Like the oracles of the ancient world, the machine didn’t provide THE answer, it simply provided plausible answers that I had to filter with my own human judgment. It’s not sentient—yet.

AI will walk with us into the future, whether we like it or not. The impact on politics and democracy will be profound. It will be helpful and unhelpful in the same breath.

It’s reminiscent of the discussions that I have with my students about the role of citizens in contemporary politics and democracies. The fog of information, misinformation and disinformation coming at our windshields is thick, but that doesn’t give us a pass as citizens when it comes to deciphering fact from fiction or truth from fallacy. Being an informed citizen is just harder than it used to be.

AI will walk with us into the future, whether we like it or not. The impact on politics and democracy will be profound. It will be helpful and unhelpful in the same breath. How? Here’s how I see the good, the bad and the ugly.

The good

Let’s start with efficiency. AI will automate tasks such as content creation, allowing strategists to focus on strategy. In the battle between those throwing stones and those having to deal with the broken windows, it will help organisations with their crisis communications by using machine learning to monitor news, identifying patterns and problems to minimise damage. It might help predict the outcomes of elections and referenda with greater precision than polls, thereby helping political campaigns make more informed decisions on where to spend their time and money. (By the way, one of the mistakes that we humans have made with polls is to convince ourselves of their accuracy; they have transparent margins-of-error, but we citizens [and politicians, governments and the media] desperately want them to reveal truths that they are not equipped to reveal. Let’s not make the same mistake with AI.)

Generative AI will take political modeling and sentiment analysis to another level. It will be deployed in political circles to personalise messages to individual voters. Texts that have been personalised based on your past voting history or the issues that you care about most will become more granular. Helpful? Slightly creepy? A logical extension of current personalisation algorithms that we currently use every day? 

AI will sift through large amounts of data to help inform government policy decisions. It could predict and help prevent instances of election interference and improve public services. It might change the balance of power between governments and citizens by giving citizens more voice and agency. It could be used within citizen assemblies to help with facilitation, information gathering, consensus building and idea generation. Citizen assemblies could be asked to recommend AI policies. AI will likely help the media in its watchdog efforts. 

The bad and the ugly

AI could give governments unprecedented surveillance power over citizens. It could exacerbate mis/disinformation and deepfakes, while simultaneously improving tools that to combat such synthetic media. AI could undermine democratic values by perpetuating and amplifying social inequalities. It could further undermine trust and the social contract (although we’ve done a sufficient job of debasing that all by ourselves). It could challenge politicians and communicators in terms of AI’s disruption of work, employment and economies.

But don’t take my word for it. Here’s what ChatGPT said when I asked it how AI will impact political communications and democracy:

“Overall, AI has the potential to significantly impact strategic and political communications, but it is important that these impacts are understood and managed appropriately to ensure that they are used ethically and effectively. The impact of AI on democracy will depend on how it is developed and deployed. Policymakers will need to carefully consider the potential risks and benefits of AI and take steps to ensure that it is used in ways that promote democratic values and protect human rights”.

- ChatGPT

We—human citizens—will need to be vigilant and involved as well, because AI isn’t something happening TO us, it’s something happening WITH us. Like every technological breakthrough in human history, it shapes us, and we shape it. In his book, Identity, Francis Fukuyama notes that identity can be used to divide or to integrate. The same can be said of AI, and we citizens have a role to play in determining the outcome.