Political Advertising in the 2024 European elections - Between Europeanisation and the protection of electoral integrity online

Introduction

As more than 370 million European citizens head to the polls this  week, political actors are turning more and more to online election campaigns. In the 2019 elections, Facebook counted over 23 million Euros in political ad spending. This year, Google has already recorded more than 45,000 individual ads posted in the EU. Given the sheer amount of time people spend online, they are increasingly consuming their news and engaging in debates on social media platforms, making these  a decisive arena for democratic discourse in the EU.

In  a  democracy,  a  fair  and  open  political  process  should  embrace transparency and accountability in political advertising. EU member states used to have vastly different rules in place to guarantee this principle in national elections. These laws, however, often focus on traditional offline media forms and are an inadequate framework for the new online arena: The risks here became evident in Trump’s 2016 presidential campaign and in the Brexit referendum where political consulting firm Cambridge Analytica used manipulative tactics to sway voters. National legislation also fails to account for the transnational nature of European elections, where cross-border campaigning requires common rules. On top of all this, keeping political debate purely within national borders is increasingly fraught within this online environment.

In response, European policymakers presented new, common rules to safeguard the integrity of European elections in 2021. The Regulation on the transparency and targeting of political advertising is intended to make political campaigns more transparent and ban extra-EU sponsorship in the immediate pre-election period. It is also supposed to remove obstacles to political advertising across the EU and, thereby, facilitate transnational campaigns online. The Regulation, however, was adopted just months before the 2024 elections because the inter-institutional negotiations dragged on. Most of its stipulations only come into effect in October 2025. Hence it is mainly the Digital Services Act (DSA) that establishes a new legal framework for online political advertising this year.

This paper describes the regulatory framework on political advertising that will hold sway in the 2024 European elections. It does so first in light of the intended goal of fostering more transnational campaigns and second, it assesses what the new rules mean for electoral integrity. While European political parties as such do some online transnational advertising, most campaigning is still carried out by national parties on national platforms. Easier cross-border advertising could pave the way for more truly European campaigns and more engagement in other member states in the future, however. The 2024 European elections will give clues as to whether the new rules suffice to ensure that online political advertising is fair and transparent. Critical here is how online platforms implement them and the transparency and access to data about the elections available on their sites.

Protecting and Europeanising the electoral processes in the EU

The regulation of election-related expression and political advertising is a sensitive matter. It involves the balancing of interests between safeguarding both freedom of expression and electoral integrity. Member states have historically landed on different solutions for this balancing act, ranging from no regulation to complete bans on political advertising during election periods. During the 2019 and previous European elections, there were no common rules for political advertising, making it difficult to conduct pan-European campaigns. European political parties for example complained about Facebook’s ban on transnational advertising. There were also no specific rules applicable to ensure that election campaigns online were fair and transparent. There was only a set of voluntary guidelines for major online platforms in the form of the Code of Practice on Disinformation. The Code was the main tool to combat undue influence, including disinformation and untransparent advertising. But this self-regulatory approach proved insufficient as platforms in the main failed to implement their commitments. The limitations of the Code became apparent – not least when X (formerly Twitter) withdrew in 2023.

The EU set out to address this deficit in time for the 2024 European elections with the regulation on the ‘transparency and targeting of political advertising’. It is intended to complement the Digital Services Act which entered into effect for large online platforms and search engines in August 2023. The DSA constituted a major overhaul of the legal environment for online platforms, including the requirement to address systemic risks to electoral processes. The regulation on political advertising, by contrast, addresses this specific issue alone and is applicable both online and offline. It is the first EU legislation directly addressing the area of election-related expression. However, the regulation only entered into force on 9 April 2024. To grant a transition period, most elements of the regulation will only apply from October 2025.

The Regulation on the Transparency and Targeting of Political Advertising

Under the regulation, political advertising will have to be clearly labelled so it can be recognised as such. It will also have to feature a transparency notice, setting out whether the ad was targeted, who sponsored it, how much the sponsor spent on their campaign, what election the ad is linked to and other details. Targeting, or the technique used to address a specific group of people based on personal data such as a user’s age bracket or location, will be curtailed. Use of personal data will only be allowed with explicit consent, sensitive data such as religion or ethnicity is banned. There will also be a ban on third country sponsored advertising three months before elections. To ensure the enforcement of these rules, the political advertising regulation stipulates a 48-hour time frame for large online platforms to process notifications about an ad in the month leading up to a vote.

Another important component is the establishment of a European repository for political advertisements designed to provide easy access to information in a user-friendly way, allowing for example for multicriteria search queries. This repository will be established and managed by the Commission and will contain copies of all online political ads for a period of seven years. Large online platforms are already obliged to create an online ad repository under the DSA. They will now have to make political ads on their platforms available in both their own and the European repository.

Most of these rules will enter into force from October 2025 but two elements are already binding: the definition of political advertising and the provisions seeking to ensure that European political actors can conduct cross-border campaigns. During the 2019 European elections,  Facebook  required  the  sponsor  of  an  advertisement  to  be  established  in the country where the campaign was supposed to run. With the new rule in place, not only European political parties but also civil society organisations and others can run transnational campaigns without having to rely on a partner in every member state.

The Digital Services Act and systemic risks to electoral processes

The DSA has also created new rules for online campaigns this year. ‘Very large online platforms and search engines’ – of which there are currently around 20 entities, including Google, Instagram, Facebook, TikTok and X – are required to assess systemic risks stemming from their services, and put adequate mitigation measures in place. One category of systemic risk is any actual or foreseeable negative impact on democratic processes, civic discourse, electoral processes and public security. These risks can relate, for example, to how advertisements are selected and presented, and the dissemination and amplification of content.

The DSA offers only guidance on what mitigation measures should look like for all systemic risks, but leaves the Commission with the option to provide guidelines on specific risks, providing platforms with more details as to what it would consider sufficient mitigation measures. On this basis, the Commission published Guidelines on the mitigation of systemic risks for electoral processes in April. It mentions the standards established through the Code of Practice on Disinformation and includes measures such as fact-checking labels, checkmarks for official accounts and options for users to control their recommender system settings, for example to reduce the prominence of disinformation. Concerning political advertising, the guidelines advise platforms to implement from the word go regulation provisions including clearly marking political ads and making sure influencers are transparent when posting sponsored political advertising.

Will the new rules ensure more European campaigns?

A key aim of the new regulation is to facilitate cross-border political advertising in the EU. Has the lifting of restrictions led to more Europeanised campaigns in the 2024 elections? Generally speaking, European political parties have become more active in the field of campaigning and advertising over the last few years. While it is difficult to gauge from Meta’s and Google’s ad repositories how much exactly parties are spending on their campaigns and how this compares to 2019, most European political are apparently engaging in transnational online campaigning: For the current campaign, the Greens are sponsoring an ad on why glyphosate is dangerous across the EU, the EPP is advocating for a stronger Frontex in Croatia and stopping organised crime in Malta. The EPP has also spent close to 80,000 Euros on a Google ad campaign promoting Ursula von der Leyen’s leadership qualities across the EU. Ironically, parts of this campaign initially failed to transparently disclose who was behind it. In comparison, the EPP spent less than 20,000 Euros on promoting then lead candidate Manfred Weber in 2019.

The harmonised rules for political advertising have made it easier for European political actors to advertise across borders this year and it appears that they have made use of this facility. Nevertheless, most of the campaigning has still been carried out by national political parties. In the absence of transnational lists and candidates, there is little incentive to spend money on European rather than national platforms that more likely capture voters‘ attention. This perpetuates the national character of the elections. In future, European political parties can and should commit to more European campaigns.

Looking ahead, easier access to cross-border online advertising could also pave the way for national political actors to get more involved in political campaigns in other member states. Facebook’s rules were initially conceived to prevent foreign meddling – the regulation on political advertising now bans third country nationals from advertising before elections but not actors within the Union. Hungary for example is known to actively advocate for preferred candidates in other countries’ elections, and be very active in the realm of online campaigning. It previously ran ads in the 2023 Slovak parliamentary election, which were not labelled as political ads and ran during the country’s silence period. A Hungarian government account also promoted ads on illegal migration in the run up to elections in Poland, Italy and Germany which were seen by millions of citizens. Once the regulation on political advertising is applied and enforced correctly, such advertising in other member states will be legal and possible without obstacles – but who is behind the advertising will become transparent. It could also allow European political parties to more easily target specific regions in the European election, for example those with historically low turnouts. This kind of involvement might just be the logical consequence of both moving online and towards a more European polity. Overall, the positive effects of greater transparency and easier cross-border campaigning are likely to outweigh any potential negative effects if and when political actors get more involved in national issues in other member states.

Will the new legal framework protect the integrity of the European elections?

When it comes to political advertising on online platforms, now predominantly on the platforms managed by Meta and Google, three main risks arise. First, technology makes it possible to ‘target’ or even ‘microtarget’ political ads. In the 2016 US Presidential election, Donald Trump for example used this strategy to deter African American citizens from voting. It can also be used to target different audiences with diverging, possibly even contradictory messages. Political parties in the EU have used this technique in the past: for example, the German FDP party was found to have shown people who are interested in ‘green’ topics that it would do more to stop climate change while promising frequent travellers it would not support any regulatory measures against climate change. Even the European Commission allegedly used microtargeted ads in September 2023 to promote a new law. Here, the Commission used micro-targeting to prevent the ads being shown to Eurosceptic users, amongst other criteria.

Using even more granular personal data, targeted political messages can potentially not only be tailored to a person’s interests but also to their character traits and vulnerabilities. Studies have shown that emotionally charged political ads that match one‘s personality traits are more likely to be persuasive. The evolution of generative AI could even make the use of personalised political messages for every individual voter feasible. The proliferation of personalised, targeted advertising and abuse of personal data presents a clear risk to transparent and fair election processes. European political parties reacted to this threat by vowing in April to no use deceptive AI generated content such as deepfakes in this election cycle.

A second issue is that of identifying and labelling political advertising: The stricter rules for political versus commercial advertising can only apply if the ad is recognised as such. Researchers found in a study across 16 EU countries that two out of three political advertisements are not declared correctly on Facebook. Facebook took action in less than five percent of cases of those unlabelled ads. Over-declaration, so a large number of advertisements that are labelled as political even though they are not, is also an issue. This can lead to inconsistent enforcement and wrongfully limit certain messages from reaching their audiences. In addition, ads that do not meet the transparency requirements need to be removed quickly. The longer wrongfully labelled ads are left on platforms, the more people would have already seen them even if they are taken down later. This has been a persistent issue in the lead up to the 2024 elections – illegal pro-Russian ads are reportedly still spreading and reaching millions of EU citizens despite having been flagged.

The third problem with political advertising is connected to the nature of social media platforms and how their feed and recommender algorithms function. Platforms are profit driven and they generate their earnings by keeping users on their site for as long as possible to show them as many ads as possible. While selecting a general target audience is up to the ad‘s sponsor, the platform has leeway in the ad‘s delivery. To maximise engagement, users are likely going to be shown content they already like and agree with. Researchers found that because of this it is actually four times more expensive to reach audiences that are likely to disagree with a certain political message. Users end up seeing only one political message rather than a plurality of viewpoints, reinforcing their pre-existing beliefs. More generally, platforms promote content that is shocking and extreme. This platform design also means that in the category of unpaid political messages, hateful political content is likely going to be shown to users more often. Researchers found that content from parties like the German far-right AfD is disseminated to users much more often than other parties’ content. This can give extremists a far bigger platform than their true level of support justifies and prevent users from seeing a plurality of viewpoints. Across the EU, far-right parties have been able to use platform logic to their advantage which could help them win votes especially amongst young Europeans in this election. They have also been spending the most on online ads in this election campaign.

Are the new rules addressing these risks? The overall higher transparency standards will be a step forward. The regulation on political advertising, once in force, limits the possibilities for targeting and applies additional transparency requirements upon targeted ads. The use of personal data is only allowed with explicit consent, sensitive data like private political views cannot be used at all. This will limit manipulative targeting techniques. Microtargeting will still be possible, in line with the GDPR. While malign actors might still find ways to circumvent the rule, the EU’s approach should make manipulative targeting less attractive and overall limit potential harm. The correct identification and labelling of political advertisements are likely going to remain an issue. The regulation on political advertising requires platforms to have a notification mechanism in place that allows users to flag an incorrectly labelled political ad, and process notifications within 48 hours in the month before an election. While this will be an improvement compared to today, it relies mainly on the self-declaration of political ad sponsors and users flagging content if it is to work. The effects of recommender systems and amplification techniques on the dissemination of political content are addressed in the DSA Guidelines. This includes for example that online platforms should give users the option to increase media diversity in their feeds.

For the 2024 European elections, however, these rules have come too late. They are only starting to be implemented and enforced. The Commission has taken an active role in this process by publishing the DSA guidelines, including encouraging platforms to implement the regulation on political advertising in advance. They also performed a stress test on platforms’ election readiness in April. With regards to enforcing the DSA, the Commission has already started two proceedings against TikTok this year. In April, it opened proceedings against Facebook and Instagram because of the suspected spread of deceptive content, limited  visibility  of  political  content,  no  third-party  monitoring  tool  and  insufficient flagging mechanisms. Nevertheless, platforms have not met all such requirements. Google, for example, is currently applying a narrower definition of political advertising, potentially allowing influencing attempts to remain undetected. Overall, the EU’s new legal framework does not yet apply comprehensively and has thus failed to reach its full potential in terms of ensuring electoral integrity.

That is why it is vital to monitor this election cycle and see where risks persist. Both the DSA and the political advertising regulation create, sometimes broadly formulated, rules in a sensitive area – leaving platforms too much leeway can have serious negative consequences for democratic processes, but over-strict rules or incorrect implementation can stifle free speech or certain political views. In order to monitor implementation, enforcement and risks, platforms need to offer adequate access to data. As the European ad repository does not yet exist, Google’s and Meta’s own repositories are the main source of information for this election. The current search options and settings allow only very limited insight into advertising data. While Google allows one to view data aggregated at EU level, Meta’s ad repository only shows ads at national level. It would also be important to filter ads based on the relevant vote, for example the European election. All in all, these limited ad repositories make it difficult to glean meaningful information on political advertising in the 2024 European election.

For future elections, it is likely that online platforms and how they function will continue to evolve, potentially developing new risks. The EU’s regulatory framework on political advertising might even affect this: Platforms could decide not to offer political advertising due to high compliance costs. The comparatively new platform TikTok that is especially popular amongst younger generations officially bans political advertising. On this platform, recommender settings will play a much bigger role than transparency rules on paid political advertising.