Since the Cambridge Analytica scandal broke out in 2018, online targeted political advertising has been the subject of increasing regulatory discussions in Europe. Attempts to cast more light on political ads online began with voluntary mechanisms of industry self-regulation, such as the EU Code of Practice on Disinformation, but in the past year we have seen a shift towards proposals of legally binding instruments. In 2021, the European Commission tabled a proposal for a regulation on the transparency and targeting of political advertising, while some EU member states also try to incorporate their own solutions into ongoing reforms of national electoral laws.
EU effort to regulate political advertising
In the proposed EU regulation, “political advertising” is defined to encompass not only electoral advertising or advertising coming from political parties and candidates, but also any advertising linked to elections, referenda, and regulatory or legislative process, regardless of who sponsors it. As a result, advertisers would be required to self-declare if their ads fall under this definition. In addition, flagging mechanisms would be put in place to allow people to bring advertisements that have been misclassified – or that are deemed unlawful – to the attention of the online platform on which they were published.
Furthermore, all political ads would be made publicly available in an online ad archive and would be accompanied by a transparency notice indicating, among other information, the sponsor of the ad as well as an explanation of personal data and parameters used for targeting. On the other hand, online platforms and other advertising intermediaries would by default not be allowed to use sensitive data such as race or sexual orientation in the process of tailoring ads. However, the feasibility of enforcing such a prohibition effectively in a world of big data and algorithm-driven automated processing of such data is debatable.
A worrying example of proposed regulation: the Irish Electoral Reform Bill
One national proposal for regulation of political advertising that has recently sparked controversy is Ireland’s Electoral Reform Bill. The proposal would require online platforms to perform an extensive verification of buyers of political ads, in order to ensure that they are allowed to buy political ads in Ireland. This would entail, among other information, collecting official documents, evidence that the person or entity resides in Ireland, and a written statement confirming the source of funds used for paying for the ad. If the platform fails to prevent the publication of an ad for which there are “reasonable grounds” to conclude that documents presented by the buyer are inaccurate, it would be held criminally liable. By putting onerous verification requirements on online platforms, the Irish bill adopts a vastly different approach than the Commission’s own proposal for identifying and marking political advertising.
The European Commission has raised concerns about this proposal in light of existing EU law. Under EU’s E-Commerce Directive, to be held liable for illegal content, online platforms should have “actual”, rather than potential knowledge of its illegal nature, and should not be required to systematically monitor all uploaded content in order to find infringements. According to the Commission, Ireland’s solution lowers this established threshold for liability in a way that is incompatible with the E-Commerce Directive (and most likely also with the EU Digital Services Act that will soon replace it).
The desire to escape responsibility could lead platforms to over-remove legitimate ads, and thus adversely impact fundamental rights and freedoms, such as the freedom of expression. Civil society could also bear the cost of this controversial proposal as many of their adverts would be categorized as “political” and therefore fall under this mechanism.
How should political advertising regulation protect civic space? Our recommendations
ECNL believes that more transparency and restrictions related to profiling and targeting internet users are necessary for effective protection of the right to data protection and privacy as well as democratic debate online. Currently, online platforms collect and analyse vast amounts of personal data about every user, often without their awareness and beyond their control. Based on this information, they predict users’ characteristics and group them into demographic categories, from which advertisers can choose when targeting the ad. Advertisers’ budget is often insufficient to reach all people who meet their criteria, so in addition to advertisers’ choices, online platforms use opaque algorithms to determine who will eventually see the ad. This surveillance-based advertising process has been shown to lead to discrimination based on gender and race, and to exploitation of vulnerabilities, which is especially concerning when private information is used to target political or socially relevant messages.
On top of that, such process can also have a negative impact on advocacy groups who want to reach new supporters: researchers have found that because algorithms are designed to maximise engagement, it is more expensive to advertise to people who, according to the platform, do not already support the advertised message.
Finally, to secure a constant influx of data for even more detailed profiling, platforms choose to prioritise attention-grabbing content which ultimately negatively impacts the quality of online public debate.
To address these problems, together with the European Partnership for Democracy (EPD), we call on EU policymakers to strengthen privacy safeguards and prevent manipulation and exploitation of vulnerabilities by:
- allowing advertisers to choose targeting criteria only based on information explicitly provided by users for advertising purposes;
- prohibiting online platforms from using any additional personal data when determining recipients of the ad.
The work on the regulation is still ongoing, but amendments presented by the rapporteur in the LIBE Committee of the European Parliament indicate support for our proposals.
Furthermore, we are concerned that placing advocacy messages by civil society in the same category as ads by actors pursuing political goals might lead to developing practices on the national level which would adversely impact freedom of association and civil society’s ability to engage in advocacy. We have already seen examples of such efforts in the above-mentioned proposal in Ireland and also in Hungary, where additional obligations or restrictions were imposed on NGOs under the pretext of engaging in “political activity”. Thus, limiting their legitimate human rights-focused advocacy and providing opportunities for disinformation campaigns about civil society.
This is why we recommend that the proposed EU regulation, while maintaining transparency rules and restrictions for the use of personal data, should draw a distinction between political advertising and advertising about socially relevant issues by public-interest organisations.