1. A Meeting of Minds in Colombo: Confronting the Invisible Forces Shaping Elections
On a late October morning in Colombo, a quiet but critical conversation unfolded about the future of democracy itself. Representatives from the Election Commission of Sri Lanka, the media, and civil society gathered under the banner of “Beyond the Algorithm,” an event organised by Democracy Reporting International (DRI) with the co-funding of the European Union. The title was deliberate. It suggested that the greatest threats to democratic elections are no longer just the loud, obvious lies told by political rivals, but the quiet, automated systems that decide what we see, what we believe, and ultimately, how we vote. DRI Executive Director Nicola Schmidt set the tone by reminding everyone that disinformation is not a uniquely Sri Lankan problem, but a global one, supercharged by the rapid rise of generative artificial intelligence. Deepfakes, manipulated voices, and fabricated images now spread at lightning speed across social media platforms, and the algorithms that govern these platforms are not neutral bystanders. They are, in many ways, the engines of this chaos. Schmidt’s message was both urgent and simple: to protect digital rights and hold tech giants accountable, we must look beyond the algorithm itself. We must examine the incentives that reward outrage, the opaque systems that amplify falsehoods, and the human consequences that follow. The room was filled with people who understood that this was not an academic exercise. They were there because they had seen the damage that disinformation can do to trust, to communities, and to the very idea that an election can be free and fair.
2. The Human Citizen Behind Every Screen: A Plea for Informed Judgement
One of the most powerful moments of the event came from A.M. Faaiz, a Member of the Election Commission of Sri Lanka. He spoke not as a technologist, but as a guardian of democratic values. His central observation was poetic and profound: algorithms decide what appears on our screens, but algorithms do not vote. Behind every screen, he reminded the audience, sits a citizen. Behind every ballot lies a judgement—a judgement formed from what that citizen has read, heard, and come to believe. This is the heart of democracy’s promise: that citizens are free to choose, and that their choice is an informed one. Faaiz did not shy away from the scale of the challenge. He acknowledged that the information environment has become polluted, that lies and half-truths travel faster than facts, and that the very conditions under which citizens form their opinions are increasingly suspect. But his response was not to call for censorship or heavy-handed control. Instead, he announced that the Election Commission, with the support of DRI and the fact-checking organisation Factum, is formulating a strategic action plan on disinformation and harmful content. They are mapping their digital capacities and shaping a roadmap for the Commission’s own digital presence. What began as a project, he said, is becoming part of the institution itself. This is a crucial shift: the guardians of elections are recognising that they must evolve to meet the threats of the digital age. Yet Faaiz was careful to articulate a clear boundary. The purpose of these efforts is to protect voters, not to police opinion. Every measure, he insisted, must remain consistent with freedom of expression and media freedom. Protecting electoral integrity must never become a justification for silencing legitimate political debate or criticism—even criticism of the Commission itself. This was a delicate balance, and he placed it at the centre of the discussion.
3. What the Research Revealed: Impersonation, Harassment, and Silent Election Lies
The event was anchored in hard data, not just good intentions. Digital Democracy Research Officer Camila Weinmann and Digital Research Analyst Prihesh Ratnayake presented findings from three recently concluded elections in Sri Lanka, and the results were sobering. Their research uncovered 66 accounts on TikTok that impersonated political parties and candidates. These were not harmless fan accounts; they were designed to deceive, to sow confusion, and to manipulate public perception. Of these, 45.5% had been active during the 2024 Presidential Election, and 21.2% of them remained active all the way until the 2025 Local Council Election. This suggests that the problem is not just a fleeting burst of election-time mischief, but a persistent, evolving threat. There was, however, a small reason for hope: following reports made to TikTok and the Election Commission, 51 of these accounts were ultimately removed. But the research revealed deeper, more troubling patterns. Women politicians faced significantly higher levels of technology-facilitated gender-based violence during the 2024–2025 election cycle. They were not just being criticised; they were being targeted, harassed, and intimidated in ways designed to drive them out of public life. The research also found that chatbots—automated artificial intelligence systems—were spreading inaccurate electoral information to voters seeking guidance. Imagine a citizen asking a simple question about how to vote, and being given false or misleading instructions by a machine that cannot tell the difference between truth and fiction. Furthermore, political advertisements were found to be running straight through the legally mandated silence period during the Parliamentary Elections campaign, a direct violation of election law. And throughout all of this, the shadow of generative AI loomed, presenting entirely new risks to electoral integrity that traditional safeguards were never designed to handle.
4. The Election Information Ecosystem: Accountability, Polarisation, and the Fight for Trust
The first panel discussion, titled “The Election Information Ecosystem,” brought together voices from the media, the Election Commission, advocacy groups, and civil society to wrestle with questions of accountability and challenge. The conversation was frank and practical. Panellists discussed the protocols that traditional newsrooms should follow when public figures make unverified claims. In an age where every politician has a social media account and every statement is broadcast instantly, how do journalists uphold the responsibility to verify without becoming the enemy of speed? There was no easy answer, but there was a clear sense that newsrooms must be more disciplined, more transparent, and more willing to slow down the cycle of misinformation. The panel also explored the effectiveness of counter-narratives, delivered through creative formats like drama, podcasts, and interviews. Facts alone, it seems, may not be enough. To combat disinformation, one must also offer compelling stories, human emotions, and cultural relevance. The discussion then turned to the relationship between the Election Commission and tech platforms. Are voluntary codes of conduct and Memorandums of Understanding effective? Or do they simply serve as public relations exercises for platforms that continue to profit from division? The panellists grappled with the need for legislative backing—the idea that these agreements must be backed by enforceable law if they are to have any real meaning. But perhaps the most sobering reflection came at the end, when rising polarisation was identified as a main threat to information integrity. Disinformation thrives in a society that is already divided. It feeds on anger and mistrust, and it drives people further apart. Without addressing the underlying social fractures, no algorithm change or content moderation policy will ever be enough.
5. Resilience in Practice: The Gruelling Work of Fact-Checking and Media Literacy
The second panel discussion, “Resilience in Practice,” took the conversation from the macro-level to the ground-level, drawing on the lived experiences of fact-checkers, advocates, and media literacy educators. The panellists did not sugar-coat the challenges they face. They spoke about the immense difficulty of verifying information in real time, especially when it emerges from obscure sources or appears in rapidly changing formats. They described the frustrating and often opaque responses they receive from social media companies during the fact-checking process. Sometimes responses come too late to matter; sometimes they do not come at all. The sense of talking to a wall, to an automated system that was never designed to serve the public interest, was palpable. Yet there was also a story of capacity-building. Local journalists are being trained to counter disinformation, not just as individuals, but as an organised front. They are learning how to trace the origins of false narratives, how to publish corrections that actually reach the same audiences as the original lies, and how to protect themselves from the harassment that often follows their work. But one issue stood above all others: a lack of interest towards funding fact-checking initiatives. Fact-checking is slow, expensive, and unglamorous. It does not produce viral moments. It does not attract large audiences. But it is an essential pillar of any functional information ecosystem. The panellists made a powerful plea for donors, philanthropists, and governments to recognise that fact-checking is not a luxury—it is a public good that must be sustained over time, not just in the weeks before an election. Closing the event, DRI Project Manager Aingkaran Kugathasan expressed heartfelt thanks to the participants and partners for their contribution to the project. His tone was one of cautious optimism: the work is hard, the obstacles are many, but the people in the room had shown that change is possible.
6. A Shared Future: Building Resilience Beyond the Event
As the event drew to a close, it was clear that “Beyond the Algorithm” was more than just a conference title. It was a call to action for a new kind of democratic resilience. The project, known as “Strengthening Resilience Against Disinformation in Sri Lanka,” has already produced eight research reports in partnership with Factum, all of which are available for the public to access at the DRI website. These reports are not meant to sit on a shelf; they are tools for change. They provide the evidence base for policy decisions, the foundation for advocacy campaigns, and the starting point for future collaboration. The conversation in Colombo did not offer any easy solutions, but it did something equally important: it brought together the key actors, it acknowledged the reality of the threat, and it committed to a path forward. The key takeaway from the research—those 66 impersonating accounts, the persistent harassment of women politicians, the silent campaign of misinformation—must serve as a warning and a guide. The 51 accounts that were removed show that accountability is possible when institutions and civil society work together. But the 15 that remained, and the countless other unseen threats, remind us that vigilance cannot be a seasonal effort. Democracy is a living system, and it requires constant care. The algorithms may be beyond our control, but the choices we make as citizens, journalists, election officials, and human beings are not. The event in Colombo was a small but significant step toward a future where the information we consume is trustworthy, where the judgement behind every ballot is genuinely informed, and where democracy’s promise is kept.