A Simple Example: How Digital Manipulation Undermines Democracy

Just a few days after launching the Mike Koedinger Foundation at Mudam, I found myself at a friend's birthday party having yet another conversation about populism, democracy and media literacy. (Incidentally, the friend I was talking to had just become one of the Foundation's very first donors.)

Once again, I was explaining how this journey started for me almost ten years ago. Around 2016, the world saw the election of Donald Trump, the United Kingdom voted for Brexit, and Germany witnessed the rapid rise of the AfD. At the same time, digital platforms were becoming the new public square. Those developments were not isolated. They were deeply interconnected.

As these conversations can easily last for hours, and I didn’t want to bother my friend lecturing him at a social event, I decided instead to illustrate my point with one simple example from a decade ago.

In 2017, Bloomberg published an excellent investigation into Harris Media, a Texas-based political marketing agency that worked with conservative campaigns around the world. The case study described how digital advertising techniques originally developed for commercial marketing were applied to political campaigns during the rise of the AfD in Germany.

The strategy leading to mass manipulation of voters can be summarised in three simple steps.

First, build a behavioural profile. By analysing the behaviour of a AFD’s existing supporters on social media, platforms create a remarkably detailed picture of who they are. Using billions of digital signals, they infer what these people care about, what they fear, what makes them angry and what motivates them to engage. This becomes the reference profile.

Second, identify lookalike audiences. Sophisticated algorithms then compare this profile with millions of other users. They identify people who behave and think in similar ways, even if they have never shown any interest in the party. These highly segmented groups become the targets of micro-targeting campaigns, enabling political messages to be delivered with extraordinary precision.

Third, personalise the message. Each audience receives content specifically designed to trigger an emotional reaction. Women might be shown fabricated stories about asyle seekers attacking women. Small business owners might receive false claims about asyle seekers taking jobs or undermining local businesses. (Remember we are in 2016, when Angela Merkel let over 1 millions asyle seekers enter Germany since 2015). The stories themselves are entirely false, but they are carefully matched to the concerns of each audience. The audience will share the messages to friends, relatives or trusted contacts, and as they gained credibility, they will spread organically across social networks and are amplified by platform algorithms. As a result, misinformation can quickly become part of people's perceived reality.

This leads to electoral manipulation as people’s voting intentions have been manipulated by fake news addressed to them based on their behaviroul profile, and amplified by algorithms creating value for shareholders while manipulating the users. And, in this case, manipulating democracy itself.

When I finished my explanation, my friend looked at me and simply asked:

"Why has nobody ever explained it this way to me?"

I've been asking myself the same question for nearly a decade.

That is precisely why I created the Mike Koedinger Foundation: to contribute to media literacy, strengthen democratic resilience and make complex subjects understandable through simple, accessible explanations.

Because understanding how our information ecosystem works is no longer a matter of curiosity. It has become a prerequisite for protecting democracy itself.