Open letter to Meta and Google regarding their decision to stop political advertising

Photo crédit : Liberties

Joint Civil Society Statement on Meta’s Decision to Stop Serving Political Ads in Europe

25 September 2025

We, the undersigned civil society organisations dedicated to protecting free and fair elections, democratic discourse, and transparency across Europe, write to articulate our deep concerns regarding Google’s recent announcement that it will update its Political Content policy in September 2025. This decision comes just ahead of the entry into application of the new Regulation on the Targeting and Transparency of Political Advertising (TTPA) in October 2025. While Google had already announced in November 2024 that it would, with limited exemptions, stop serving political advertisements in the European Union, the new announcement specifies in more detail which types of ads will be banned and which may remain permitted.

In December 2024, some of us already voiced concerns about Google’s withdrawal. At that time, one of our major concerns was that this would leave Meta with a de facto monopoly in the online political advertising market, with potentially harmful consequences for democratic debate. With Meta also now exiting the market, that specific scenario will not materialise. However, the combined withdrawal of the two largest platforms that previously carried political ads effectively leaves the EU without any major online platform offering political advertising. This marks a significant shift in the digital public sphere.

In this context, we remind Google that under Article 34(1)(c) of the Digital Services Act (DSA), Very Large Online Platforms and Search Engines, including Google’s relevant services, have the obligation to carry out risk assessments regarding any actual or foreseeable negative effects on civic discourse and electoral processes. Under Article 35, they are required to put in place reasonable, proportionate, and effective mitigation measures.

Engagement-optimising algorithms are designed to maximise user attention, which in practice means privileging emotionally charged content. In such an environment, political advertising can serve as one of the few channels through which moderate actors reach their audiences. Unless Google makes corresponding changes to its recommender systems, the new ban on political ads risks reducing the visibility of moderate voices and thereby accelerating radicalisation, to the further detriment of democratic discourse.

It is to be noted that upon reading the announcement, we do believe that the scope of the restrictions is not sufficiently clear. We are concerned that, to minimise compliance costs and the risk of fines, along with sui generis political ads, Google, in practice, will also block ads from civil society organisations drawing attention to societal issues or calling for action. This would severely limit the visibility of CSOs conducting important advocacy and awareness campaigns, and would negatively impact their fundraising activities, which are essential for their survival.

It should also be noted that while Google’s political ad offer was far from perfect, including shortcomings with data quality and user-unfriendly ad repositories, it provided a more desirable model than Meta’s. With further restrictions on targeting options, Google could still play a constructive role in supporting pluralistic democratic debate. If Google is unwilling to reconsider the ban on political advertising, then at the very least, it should commit to transparency regarding the effects of withdrawing political advertising on public discourse and the mitigation measures it intends to introduce

In this context, we call on Google to :
• Re-examine the possibility of serving political advertising in the EU that is not based on tracking and profiling, in a manner that is more conducive to civic discourse and electoral processes.
• Ensure that civil society organisations and other actors can continue to run legitimate civic engagement campaigns, including fundraising and issue advocacy, without arbitrary restriction.
• Disclose how its algorithmic systems affect political and civic content, including what steps have been taken to avoid systematic biases that may silence moderate voices or advantage emotionally charged, polarising content.
• Re-engage with EU institutions, civil society, and researchers to ensure that its evolving political content policies are transparent and compliant with the EU’s regulatory framework, including, but not limited to, its obligations under the Digital Services Act to mitigate risks to civic discourse and electoral processes

We welcome further dialogue with Google on these matters and urge the company to publish a clear policy statement and risk assessment outlining how it intends to support democratic discourse in the EU going forward.

– - - - - - - - - - - -

oint Civil Society Statement on Meta’s Decision to Stop Serving Political Ads in Europe

25 September 2025

We, the undersigned civil society organisations working to safeguard free and fair elections, democratic discourse, and transparency in Europe, are writing to express serious concerns about the impact of Meta’s recent decision to stop serving “political, electoral and social issue ads” in the EU ahead of the entry into application of the new Regulation on the Targeting and Transparency of Political Advertising (TTPA) in October 2025.

While the TTPA has a number of serious weaknesses, Meta’s decision to ban “political, electoral and social issue ads” while risks creating a growing imbalance in the online public sphere. Assuming that Meta will simultaneously maintain its algorithmic systems that prioritise emotionally engaging content, a broad but inconsistently moderated ban of certain types of ads may disproportionately disadvantage moderate and non-partisan political actors, who rely on reasoned argument rather than divisive speech when publishing paid content on Meta platforms.

Under Article 34(1)(c) of the Digital Services Act (DSA), Meta, along with all other Very Large Online Platforms and Search Engines, has the obligation to carry out risk assessments regarding any actual or foreseeable negative effects on civic discourse and electoral processes and, under Article 35, to put in place reasonable, proportionate, and effective mitigation measures. We believe that Meta’s efforts to effectively and demonstrably mitigate the risks its algorithmic content recommender systems pose to civic discourse and electoral processes have so far been unsatisfactory, and that this new policy change is likely to lead to a further deterioration of the situation.

There is ample evidence that similar social media platforms that have introduced such broad bans, have consistently failed at correctly delineating and filtering “political, electoral and social issue ads”. The result is that the ban legally lets the platforms off the TTPA’s hook, disregarding transparency requirements and targeting limitations, while many of the ads that aren’t supposed to be there anymore, continue to be published.

In addition, Meta’s definition of “political, electoral, and social issue ads” is very broad, and in many countries, Meta platforms’ market dominance is such that they constitute the major, or even the only, online space where moderate political actors and civil society organisations can reach people. That is why the new restrictions will severely limit those organisations’ visibility and could negatively impact their fundraising activities, which are essential for their survival.

It is important to note that we are not calling for the continuation of hyper-personalised advertising, which is currently at the core of Meta’s business model. Meta has argued that the TTPA makes it impossible to continue offering political advertising in the EU, claiming that restrictions on targeting undermine products advertisers rely on and prevent users from seeing ads it considers relevant. We believe this argument is fundamentally flawed. Hyper-personalised advertising has no place in a society striving to maintain democratic discourse.

In this context, we call on Meta to :
• Re-examine the possibility of serving political advertising in the EU that is not based on tracking and profiling, in a manner that is more conducive to civic discourse and electoral processes.
• Ensure that civil society organisations and other actors can conduct legitimate civic engagement campaigns, including fundraising and issue advocacy, without arbitrary restriction.
• Disclose how its engagement-based ranking systems affect political and civic content, including what steps have been taken to avoid systematic biases that may silence moderate voices or advantage emotionally charged, polarising content.
• Re-engage with EU institutions, civil society, and researchers to ensure that its evolving political content policies are transparent and compliant with the EU’s regulatory framework, including but not limited to its obligation under the Digital Services Act to mitigate risks to civic discourse and electoral processes.

We welcome further dialogue with Meta on these matters and urge the company to publish a clear
policy statement and risk assessment outlining how it intends to support democratic discourse in the
EU going forward