Global Elections and Platform Responsibility
This year is a major one for elections around the world, with pivotal races in the U.S., the UK, the European Union, Russia, and India, to name just a few. Social media platforms play a crucial role in democratic engagement by enabling users to participate in public discourse and by providing access to information, especially as public figures increasingly engage with voters directly. Unfortunately elections also attract a sometimes dangerous amount of disinformation, filling users’ news feed with ads touting conspiracy theories about candidates, false news stories about stolen elections, and so on.
Online election disinformation and misinformation can have real world consequences in the U.S. and all over the world. The EU Commission and other regulators are therefore formulating measures platforms could take to address disinformation related to elections.
Given their dominance over the online information space, providers of Very Large Online Platforms (VLOPs), as sites with over 45 million users in the EU are called, have unique power to influence outcomes. Platforms are driven by economic incentives that may not align with democratic values, and that disconnect may be embedded in the design of their systems. For example, features like engagement-driven recommender systems may prioritize and amplify disinformation, divisive content, and incitement to violence. That effect, combined with a significant lack of transparency and targeting techniques, can too easily undermine free, fair, and well-informed electoral processes.
Digital Services Act and EU Commission Guidelines
The EU Digital Services Act (DSA) contains a set of sweeping regulations about online-content governance and responsibility for digital services that make X, Facebook, and other platforms subject in many ways to the European Commission and national authorities. It focuses on content moderation processes on platforms, limits targeted ads, and enhances transparency for users. However, the DSA also grants considerable power to authorities to flag content and investigate anonymous users – powers that they may be tempted to mis-use with elections looming. The DSA also obliges VLOPs to assess and mitigate systemic risks, but it is unclear what those obligations mean in practice. Much will depend on how social media platforms interpret their obligations under the DSA, and how European Union authorities enforce the regulation.
We therefore support the initiative by the EU Commission to gather views about what measures the Commission should call on platforms to take to mitigate specific risks linked to disinformation and electoral processes.
Together with ARTICLE 19, we have submitted comments to the EU Commission on future guidelines for platforms. In our response, we recommend that the guidelines prioritize best practices, instead of policing speech. Furthermore, DSA risk assessment and mitigation compliance evaluations should focus primarily on ensuring respect for fundamental rights.
We further argue against using watermarking of AI content to curb disinformation, and caution against the draft guidelines’ broadly phrased recommendation that platforms should exchange information with national authorities. Any such exchanges should take care to respect human rights, beginning with a transparent process. We also recommend that the guidelines pay particular attention to attacks against minority groups or online harassment and abuse of female candidates, lest such attacks further silence those parts of the population who are already often denied a voice.
EFF and ARTICLE 19 Submission: https://www.eff.org/document/joint-submission-euelections
