Reflections from Berlin: How AI could bolster democracy

In the very spirit of democracy, a group of leaders, changemakers, and innovators came together in Berlin on 17 April to discuss how democratic culture could be strengthened in the era of AI. A major takeaway from ​​“​The Civic Toolkit: Digital Literacy in the Era of AI,” organised by The Conduit Foundation, was that AI could be a powerful tool for good, but only if it is in the hands of people who want to instigate positive change.  

The diverse group came with their own personal, diverse experience – both professional and personal: some had a background in tech and policy, others arts and education, and while some grew up with the internet and social media as omnipresent, others remembered the time when the internet heralded a democratisation of information for all. The common factor between the group: everybody was there to collaborate on building bold solutions that matter for the future of democratic culture.   

The event took place less than a week after the historic defeat of Victor Orban’s regime in Hungary, and while that was a cause for some optimism, many in the audience felt the urgency of the threats to the very principles of the democratic system. As Paul van Zyl, CEO of The Conduit and Trustee of The Conduit Foundation, noted in his opening address, only 21% of the world currently lives in a country described as free, down from 46% twenty years ago. “We can’t achieve the system change that the world needs by going it alone,” he said. 

While being honest about this downward trend in these tumultuous times, speakers also urged participants to seize the opportunity. As Olivia Lazard, Planetary Security & Ecology Expert at the Berggruen Institute, said in her opening keynote, “the moment that things break down is the moment we have breakthroughs.” 

AI can be a useful tool, but the politics currently driving it are problematic

One issue very much front of mind for the audience was the link between AI and how it’s being used in certain political circles. The example was given of President Trump’s spat with Pope Leo, followed by the President sharing on social media an AI-generated image of himself as Jesus (or a doctor as he claims he thought it was). Much of the criticism of AI at the event was levied at powerful people using it for nefarious means. “Trump is one example of how AI has been used to build personality cults,” said Julia Ebner, Co-Executive Director of the Institute for Strategic Dialogue.

Matthias Spielkamp, Executive Director of Algorithm Watch, said during a panel that AI in its current form is being driven by fascists and authoritarians. “It’s not the technology, it’s the people behind it who can be a problem for democracy,” he said. Ramak Molavi Vasse’i, Head of Advocacy at the Center for Digital Rights and Democracy, told the panel that we must be careful of the potential shift in power that the swift deployment of AI could bring. “As long as we don’t tackle the ownership and infrastructure, this will not magically turn into a good thing,” she said. 

Similar sentiments were threaded throughout the day, and attendees had the opportunity to explore them further through “Democracy  Lab” deep dives into the overlap between AI and core democratic values like resilience, inclusion, and dialogue.   

Participants talked about how there needs to be more accountability for generative AI tools, even if they didn’t  completely agree what that should look like. Others spoke about how they would have more trust in AI if there were better controls in place, that regulation needs to come from outside the companies pushing this technology, and how AI could potentially warp democratic debate in similar ways to what we’ve seen with social media.   

AI presents an opportunity to bring democracy into the ​​21st century  

Despite strong criticisms of political misuses of AI, there was much enthusiasm for the role the technology itself could play in updating democracy. At one point, an audience member noted that AI has helped us to break down language barriers, for example, meaning people with different language skills can now build connections with one another. There was a pervading sense that AI is changing our democracies whether we want it to or not, and we need to act accordingly.  

Dimitris Dimitriadis, CEO of TheFutureCats, talked about how AI can help us fill in some of the inefficiencies in how our institutions adapt to technology. “We need to bridge the gap between how fast technology moves and how fast our institutions learn,” he said. That idea was reflected in a number of the conversations throughout the day. Robin Frasch spoke about his work on wahl.chat, a non-profit civic tech initiative that develops AI-powered tools to make politics more accessible and easier to understand. The app uses AI to take complex political content that would require a large time investment to understand, and turns it into clear, user-friendly explanations. 

Niclas Böhmer of the Hasso Plattner Institute looked at why democracy currently feels “outdated” – the world is a very different place than it was in the 18th century during the democratic revival of the  Enlightenment. Not least of all, there was a massive population explosion in the 20th century, meaning there are now many more voices in the room. While a human would find it almost impossible to mediate 10,000 voices, AI could find structured signals from within the chaos of all those opinions.​​​  

Till Behnke, Co-Founder and CEO of the Rulemapping Group, spoke about how digitalisation has completely changed the world, as well as how we record and share information. Our legal systems are currently analogue, meaning laws are published as physical texts that then must​ be translated into code in a complex process that takes time. He argued that AI could help make that a much quicker process to develop machine-readable code. ​ 

During the Democracy Labs, one participant suggested that AI could help older people, who may lack digital literacy  relative to other age cohorts, develop digital skills. Another idea floated during the discussions was to stagger access to AI tools based on completion of digital literacy modules — akin to receiving a driving licence.   

Taking AI back into the hands of those who want to do good in the world  

Regardless of where people fell on the topic of AI, it was clear by the close of the day that people wanted more to be done to protect our democracies. For AI to have a truly positive role, however, we need an urgent, firm response to the current political  – and other – misuse of it. “The early adopters are malevolent, but the late adopters can be benevolent,” said van Zyl in his closing address. “I refuse to let the early adopters set the pace.”   

​​The day’s discussions coalesced around a number of points for future reflection and action. Participants pushed back on the idea that AI was either the cause of democratic of decline or the solution to fix it. Instead they saw it as one tool in a wider repository.​  

​​There was universal agreement that relying on technology alone won’t bring an easy fix to the problems ailing democracy. One of the groups in the Democracy Lab argued that periodically going offline and ensuring there were more “third spaces” like libraries and recreational centres that offer deliberate proximity to other groups of people who we might not otherwise meet. Some friction, they argued, is necessary for our societies to function and remind us how to get along.  

When using AI specifically to defend democracy, more needs to be done by our institutions, so that the people in charge of deploying and regulating it actually understand it. While some participants called for less but better regulation in the EU, others wanted more regulation. The underlying point being ​that people ​don’t think the current legislation is up to the task, and certainly isn’t responding fast enough to ongoing changes.   

​​​It therefore falls to those who are enthusiastic about the role AI can play in shaping and bolstering democracy for the better to shift the narrative around the technology – through action. This means assuaging the concerns people have by practically demonstrating how AI can be used for good, leverage funding to get initiatives off the ground that point technology in the right direction and serve humanity and reinforce democratic culture.  

The Civic Toolkit: Digital Literacy in the Era of AI was the first in a series of three events for The Conduit Foundation in Berlin and aims to uncover the synergies to build coalitions and leverage this tool and build stronger democracies in a world of declining trust. 

All photos by Studio Kroll