AI is a powerful tool under the right democratic framework

The ancient Greek concept of Pharmakon, that something can be either a remedy or poison, sums up the potential of AI in times of democratic crises. Former TED curator Bruno Giussani introduced the idea on September 30th, when The Conduit Foundation convened a cohort of leaders, changemakers, and innovators in Berlin for the second event in a series on democratic culture and AI.

While many opposing viewpoints surfaced during the dynamic event, there was an overwhelming consensus that democratic culture needed to be revamped for the modern world.

As attendees arrived, they were asked to place a marker on a board that summed up their feelings on the ability of AI to improve democratic culture: optimistic, sceptical, or pessimistic. A majority of people stood on the optimistic side, followed closely by those who were sceptical, with the pessimists making up a small minority.

The shadow of autocracy is growing

In his opening address, Giussani set out the specifics of AI as our modern day Pharmakon. He spoke of how it’s becoming increasingly difficult to find reality amidst a wave of disinformation and AI-generated content, particularly given the way that big tech companies operate. “The way these technologies are developed are fundamentally authoritarian,” he said. “While elected governments feel, at best, powerless if not complacent.”

Photo credit: Studio Kroll

Nonetheless, Giussani also spoke about how the effects of AI on democracy will be determined by the choices we make as humans today. “Digital technology is no longer a tool we use.. it has become our environment… we swim in it,” he said. “No technology once invented has been uninvented.”

AI is too powerful a tool to ignore

To properly determine the role that AI could play in European democracy in the coming decade, a debate took place across three topics: disinformation and civic challenge, youth participation and civic agency, and hyperlocalisation and democratic renewal. Each topic had dedicated speakers to speak for and against the motion of the day: this house believes AI will do more for EU democracy in the next decade than any institutional reform.

Before the debate started, a show of hands revealed that most of the audience was against the motion. As the pro speaker on the disinformation topic, AI entrepreneur Maxim Nitsche stated that he was “happy to be a rebel” and on the pro side. He argued that while, yes, nefarious actors are using AI to spread disinformation, AI itself has also made it easier to track where disinformation is coming from — and therefore fight it more effectively.

Paul van Zyl, CEO of The Conduit and trustee of The Conduit Foundation, argued that when muskets were developed, those wielding them won battles versus those who still used spears. The implication was that AI is a powerful tool, and we will lose the fight against disinformation if we don’t use it.

Photo credit: Studio Kroll

An audience member shared similar sentiments: “We’re fiddling with pistols while others have built nuclear weapons,” he said, arguing that we need a big intervention — not the slowness that would come with institutional reform. “We’re in a different game right now and we need urgency.”

The dangers of building democracy on tools we don’t control

When AI entrepreneur Ina Remmers took to the floor as the speaker against the motion on the topic of hyperlocalisation and democratic renewal during the debate, she had a different perspective. She argued that AI is only as democratic as the institutions governing it. Referring back to the Pharmakon concept, she said: “Right now in Europe, we’re giving out the poison first and writing the label later.”

Another audience member concurred: “Even if we use AI in service of citizens… if it’s built on undemocratic infrastructure, it won’t lead us to a more democratic future.”

Electronic election researcher Joe Litobarski talked about how AI couldn’t be detached from its geopolitical reality — especially in Europe. AI is being developed outside of Europe’s democratic control, and building reliance on it could lead to problems down the road.

In May last year, for example, the chief prosecutor of the International Criminal Court (ICC) lost access to his Microsoft email after sanctions imposed by Trump. “If we become dependent on [technologies], they can be switched off,” said Litobarski.

In her closing keynote for the day, Josephine Ballon tackled the same subject. Her organisation, HateAid, received travel bans to the US and threats of economic sanctions for fighting for democratic accountability on big tech platforms.

Photo credit: Studio Kroll

While the idea that Ballon and her colleagues had been branded agents of the “global censorship-industrial complex” added some humour to the day, the attack had real life consequences. “People now think twice if they want to be in a photo with us, collaborate with us, or sponsor us,” she said.

Discussions on disinformation, agency, and participation

Later in the day, attendees broke off into smaller groups for working studio discussions to dive into the topics highlighted in the debate. A group focused on youth participation and civic agency talked about how elections are now being fought on social media, and AI could be a tool for writing scripts and helping to get videos out there. Another group focused on hyperlocalisation and democratic renewal suggested AI could find the common issues people care about, so politics is more focused on issues than party loyalty.

One person spoke about how the people consuming disinformation likely didn’t care — that they wouldn’t change their mind even if AI could provide correct information. Another in the same group raised her concerns that disinformation was only talked about in the context of authoritarian regimes, when it’s a much broader problem.

To report back on their findings and recommendations, participants were invited to join a ‘fishbowl’ — a dynamic conversation circle where anyone could sit in and voice their thoughts. On the topic of youth participation, one point came up that had also emerged in the studio sessions: there were not enough young people in the room. “We are not youth,” said one participant. “Once we understand their hopes and fears, we can deal with them better.”

Solutions for a more democratic future

Despite the many different viewpoints, several common solutions emerged from the day’s discussions on how to strengthen democratic culture in times of AI. There was a large focus on more local contexts, such as engaging in more democratic behaviours like showing up for your neighbours, smaller projects that invite specific audiences to exchange information in a casual environment such as group meals or book clubs, and dedicated public spaces where young people can go to come up with their own projects related to AI.

“Meeting people where they are even if they’re distanced from democratic processes is good for increasing participation,” said an audience member.

Photo Credit: Studio Kroll

Throughout the day, there was also a sense that AI alone wasn’t enough to solve our democratic woes. A proposed idea was “democracy duty,” where everybody has an obligation to engage in decision making at a local level — similar to jury duty. “Democracy is a contact sport, you need to get dirty,” said one of the group facilitators.

The event was the second of three on the topic of democratic culture in times of AI. Where the first focused on discovering the trends and topics at the forefront of people’s minds, the second focused more on potential solutions.

The final event will be in Berlin on November 4th, exploring how AI and digital tools can put citizens at the heart of local democracy. You can find further information on the event page and sign up to participate.

By Jack McGovan