Protection vs. censorship: how to strike the right balance curbing online disinformation

The days when conspiracy theories, disinformation and fake news were confined to the internet’s fringes are long gone. They are now part of mainstream global debate, influencing elections, disrupting markets and challenging the boundaries of free speech. As the World Economic Forum’s Global Risks Report 2025 revealed, disinformation remains the top short-term risk for the second consecutive year. It has the power to erode trust, deepen divisions, and weaken governance and global cooperation. With regulation still lagging behind the fast pace of technological change, the risk is not only persistent but growing.

At the World Economic Forum’s Annual Meeting this year, two key sessions – Truth vs. Myth in Elections and To Moderate or Not to Moderate? – explored the growing challenges of disinformation, the dilemmas of content moderation, and the fragile balance between democratic integrity, human rights and technological innovation. While panellists agreed that the current trajectory is unsustainable, the question is whether consensus can be forged amid the complexities faced by digital platforms: balancing their responsibility to curb harmful content, without undermining free expression or further polarizing societies.

In 2024, 50% of the global population went to the polls, with false narratives and misleading information shaping political landscapes worldwide, deepening geopolitical fragmentation.

The case of Moldova is illustrative. Its Prime Minister, Dorin Recean, described how foreign disinformation networks interfered in his country’s elections by disseminating AI-generated deepfakes showing children in military uniforms beside EU flags, falsely claiming that EU integration would lead to war. Disinformation is not just about lies. It is about fear, emotions and division. “They are trying to associate EU with war and say, ‘Look what happens in Ukraine. If you want the EU, the same will happen in Moldova,’” Recean warned, adding that 2.5% of Moldova’s GDP was invested in disinformation attacks.

As governments begin to understand the implications of disinformation’s impact, the European Commission’s Digital Services Act (DSA) is emerging as the benchmark in discussions on accountability. “What is illegal offline must be illegal online,” declared Jean-Noël Barrot, France’s Minister for Europe. “Transparency in algorithms is non-negotiable,” he argued. “In France, racist or homophobic remarks are against the law and can be judged by a court, so platforms should follow the same rules online. When it comes to misinformation, the DSA requires platforms to take steps to reduce the systemic risks that goes with disinformation. This is becoming increasingly essential, especially given how younger generations engage with information.”

Yet Sasha Havlicek of the Institute for Strategic Dialogue cautioned against oversimplification and warned that bad actors exploit openness on digital platforms, using deceptive tactics like botnets and fake accounts to distort public discourse in ways invisible to the average user. “The algorithmic infrastructure amplifies whatever is most likely to engage you. The business model of the attention economy disproportionately amplifies the most sensationalist, the most extreme rather than the moderate,” she said.

Moderating billions of users is not an easy task. Content moderation remains technically complex and politically contentious. In 2024 alone, the top platforms moderated 1.6 billion pieces of content, according to Michael McGrath, a European Commissioner for Democracy, Justice and the Rule of Law. “Freedom of expression must be protected, and there is no question about that. It is non-negotiable. But there are guardrails to strike the right balance,” he added.

Moderation at scale comes with pitfalls. Tirana Hassan, former Executive Director of Human Rights Watch, stressed that moderation policies must be grounded in human rights principles to prevent misuse. “Moderation should protect, not silence. Over-regulation risks enabling authoritarian control,” she said. “Investing in good content moderation, both digital and human, can mean that you are meeting legal obligations. It is not in conflict,” she said. Tirana also stressed the need to protect vulnerable groups online, noting that in some cases they were baited and outed on platforms, risking physical harm and even displacement.

Volker Türk, the United Nations High Commissioner for Human Rights, cautioned that “a lot of these products end up in situations of crisis or conflict, where there is a heightened responsibility to ensure they are not used to dehumanize or fuel more violence”.

Meanwhile, tech firms face a paradox. Generative AI has amplified the spread of false and misleading content, but also holds promise for identifying and filtering it. Some of these challenges stem from how platforms are structured and algorithms that often amplify content in ways with unexpected societal consequences.

According to Havlicek, engagement-first algorithms promote controversy over content. “Because of the attention economy, it is not a free speech environment. It is a curated one, where tech giants ultimately decide what you see based on the data they collect and the money they are going to make,” she said.

And it is not just institutions at fault. According to the 2025 Edelman Trust Barometer, globally, 40% say they approve of hostile activism, including attacking people online or intentionally spreading disinformation as a legitimate tool to drive change. This sentiment is especially predominant among respondents aged 18-34, with 53% of them saying they would approve intentionally spreading disinformation for a cause they support. This alarming trend reflects a deepening crisis of trust. Individuals are now embracing tactics once considered unethical.

These two sessions in Davos provided not just an analysis of the impact of disinformation and how to moderate social media content, but also offered direction. Here are some key priorities that emerged on how to address disinformation challenges:

Journalists and civil society organizations need to play a crucial role in countering disinformation through investigative reporting, fact-checking, and public education initiatives. Latour underscored the importance of maintaining editorial independence and transparency to rebuild public trust in the media. “The suggestion that disinformation didn’t play a large role in the past year [vis à vis elections] is grotesque. That is misinformation,” he added.

In an era dominated by algorithmic amplification and synthetic content, countering disinformation demands more than just technological solutions or regulatory frameworks. It requires shared responsibility rooted in critical thinking, transparency and resilience.

Open debate is vital to democracy. Rebuilding trust is not about consensus. The fight for truth is not about “winning”. It is about fostering environments where diverse perspectives can coexist without hostility.