There is no legitimate need for parties to target people based on their interests or presumed value choices. Parties should present the same messages to us all, as previously (Photo: Eduardo Woo)
Imagine scrolling through your favourite social media platform one early May morning. You read that Party A, which you had considered voting for but have since developed strong reservations, promises to do everything within its power to ensure business interests do not override the need to protect the environment.
“Great”, you think, “Party A politicians are finally coming to their senses.” Coincidentally, your neighour, a keen investor in the stock market, is also scrolling social media that morning. Like you, he sees advertisements from Party A in his feed. The messages he receives, however, highlight the party’s intention to prioritise business interests over environmental issues. “Finally, Party A politicians are coming to their senses”, he sighs contentedly.
And in June, both of you vote for Party A in the European Parliamentary elections.
As dystopian as this scenario might sound, it is not at all far-fetched. It is inspired by a real case uncovered three years ago by the German comedy show ZDF Royale, in collaboration with UK watchdog Who Targets Me.
ZDF Royale’s research into online political ad targeting practices by German political parties during the federal election campaign revealed how parties, in relative secrecy, can target us with different, at times contradictory, messages based on what Facebook knows about us. Their research showed that political micro-targeting can threaten the integrity of our elections.
A lot has happened since then. The European Union adopted a Regulation on Targeting and Transparency of Political Advertising.
Civil rights watchdogs hope that, if enforced appropriately, it could help to halt the erosion of European democracies. However, this regulation will not offer immediate relief in the coming months, as its key provisions will not yet be in effect.
On a brighter note, the Digital Services Act (DSA), the EU regulation designed to foster a safer digital space, became fully operational on 17 February 2024. Under the DSA, so-called ‘Very Large Online Platforms’ and search engines are mandated to conduct systemic risk assessments in key areas, including their impact on electoral processes, and to put in place related mitigation measures.
Little indication is provided on how to assess the potential risks and what the possible “reasonable, proportionate and effective” risk mitigation measures are, so the effectiveness of the DSA in protecting election integrity hinges on whether online platforms are provided with clear and appropriate guidelines and whether they are willing to adhere to these standards.
In preparation for one of the most significant election years in history, which includes EU citizens voting to elect members of the European Parliament, the European Commission has recently released draft guidelines for very large online platforms and search engines on the mitigation of systemic risks for electoral processes and launched a public consultation to seek input.
The draft guidelines identify several areas where mitigation measures could be applied, one of them being political advertising. The guidelines recommend very agreeable measures, such as the clear labelling of paid political content or the maintaining of a publicly available and searchable repository of political ads.
Overlooked risks
Notably, however, the draft guidelines overlook the risks posed by political micro- and nano-targeting practices. They fail to mention how the extremely fine-grained targeting, made possible through platforms using information gathered from constantly observing our online activities, allows political actors to target swing voters and deliver tailored, and possibly contradictory, political messages to different groups of voters. Of course, no mitigation measure has been proposed.
To ensure the democratic nature of the upcoming elections, the revised version of the guidelines should propose that very large online platforms and search engines stop processing all observed and inferred data to target or deliver political advertising and significantly limit targeting options available for political advertising. There is no legitimate need for parties to target people based on their interests or presumed value choices. Parties should present the same messages to us all, as before.
Finally, the guidelines give very large online platforms and search engines suggestions about how to fulfil their legal duties but do not establish legal obligations.
This means that even if the commission were to suggest the aforementioned measures, very large online platforms and search engines might theoretically opt to adopt different measures in the area of political advertising, arguing that their selected approaches are genuinely “reasonable, proportionate and effective”.
Without legal obligations/enforcement tools no document, regardless of how well-intentioned it is, will compel very large online platforms and search engines to comply beyond the bare minimum if it requires sacrificing profit. This is why, as the commission revises the guidelines, it must emphasise the credible threat that inadequate mitigation measures will result in significant consequences.
You and your neighbour deserve to know the truth about what Party A stands for. If you were both presented with honest information, you’d most likely vote for different parties by the time June comes around. And in a healthy democracy, this is how it should be.
Source: euobserver.com
