We are HEIO, the Hybrid Election Integrity Observatory.
- Hybrid: What happens online impacts the world offline.
- Election: We primarily research democratic elections, mostly in the Netherlands.
- Integrity: Are the rules being followed? Do campaigning, debate, and elections withstand scrutiny? Are voters and candidates getting a fair chance?
- Observatories: We monitor trends, outliers, and anomalies across online platforms to identify inauthentic and coordinated manipulation efforts on social media.
We share findings in real-time through various channels and report violations to those affected, platforms, Dutch and European authorities.
The observatory consists of:
Post-X Society
Post-X Society coordinates the team and brings the various streams of research together. Often, the most interesting findings come when we see trends across various platforms or methods – that’s when we know we’re onto something.
What they hope to find: Nothing! But we will share our observations, findings, and recommendations with government agencies – how platforms respond when users or groups violate their policies or the law; or when we find politicians, organised groups, or foreign actors using social media in a way that’s contrary to the laws and requirements of our democracy. The agencies who regulate these matters cannot conduct research themselves, so they depend on people like us and our findings..
Trollrensics
Trollrensics identifies coordinated inauthentic behaviour – cooperation through liking, sharing, retweeting, or whichever techniques artificially promote the visibility of certain messages.
What they are looking out for: We are not interested in debunking. Even when you spread facts, you can manipulate the public discourse. I’m reminded of the pilot in the John Cleese sketch announcing, “The wings are NOT on fire!” Rather than engage in an argument over what is true, we focus on identifying organised networks that are normalising extreme and hateful content by artificially boosting visibility.
AI Forensics
AI Forensics is a European non-profit that investigates influential and opaque algorithms. In HEIO, they research political content recommendations (think of rabbit holes and filter bubbles), political ad moderation (do platforms actually regulate ads?), and political narratives and topics (what’s trending in the lead up to Dutch and French municipal elections?).
How they view HEIO: All of our findings are interesting separately, but together we methodically divide work and address the most significant risks facing the elections. It is not ‘oh someone found something’ – it is a systematic review. The continuity also makes a difference. The fact that there is a consortium working together election after election allows us to hone in on difficult questions, focus resources on specific relevant angles, and compare things we find this time with things we found last time.
University of Amsterdam & CampAIgn Tracker
The University of Amsterdam and CampAIgn Tracker lead two research streams in HEIO: They identify and explore the use of AI in political discourse, and they examine the impact (or lack thereof) of self-imposed bans on political advertising by Google and Meta.
On AI in political discourse: Some people say AI is destroying democracy. Other people say AI will save it. We want to understand how AI is really being used to influence politics, whether that is through the use of deepfakes to spread disinformation, or by levelling the playing field for smaller parties who traditionally lack the budget to create content. AI is new; we want to understand how it is being used, what is problematic, and what is not so problematic when it comes to democratic elections.
