How Can Tech Innovation Better Fight Against Fake News ? Greentech Meetup 2025

Climate disinformation has crossed a critical threshold. In 2024, climate-skeptic content increased by 43%, according to analyses shared by Data for Good and Science Feedback. This is no longer a fringe phenomenon: it is structured, strategic, and amplified by the very technologies designed to broaden access to information.

An Industrialized Disinformation System

Research presented by Emmanuel Vincent (Science Feedback) shows a pattern: climate disinformation is produced methodically. In eight months, 529 documented cases were identified, grouped into 19 dominant narratives. Two findings stand out :

• 90% of attacks target climate solutions, not just climate science.
• False content benefits from algorithmic overexposure.

On YouTube, accounts spreading disinformation receive 8x more visibility. Across Europe (4 countries, 6 platforms):

• TikTok: 20% of climate content contains disinformation
• Instagram: 15%
• LinkedIn: 2%

This is a structural issue, not an incidental one. QuotaClimat and Science Feedback partnered to run a study on disinformation : read the detailed report here.

Where the Narratives Come From

These narratives are not spontaneous. They are engineered. Think tanks such as IFRAP in France and the Heartland Institute in the US produce talking points that are then amplified by political actors with direct economic interests. The US already uses advanced scientific methods to trace the origins of these narratives. France and Brazil are now experimenting with cross-border tracing frameworks, according to Eva Morel (Quota Climat).

Technology as Both Weapon and Shield

Camille Grenier (Forum on Information and Democracy) highlighted a core problem: those who control the technology are not necessarily aligned with the fight against disinformation. Platforms have become active vectors of disinformation through their algorithms. At the same time, defensive tools are emerging:

• Specialized LLM models trained to detect false information
• Models trained directly on disinformation datasets
• Systematic human validation for complex cases
• The AI tool “Spinoza” developed by the association Reporters Without Borders for journalists
• ClimateQ&A.com, built on IPCC data

 However, some tools are misused or misinterpreted and can paradoxically contribute to confusion between weather and climate.

Sora 2 and the Deepfake Turning Point

Hyper-realistic video generators such as Sora 2 (OpenAI) represent a major shift: disinformation is no longer only textual or statistical, but visual and emotional. The boundary between real and synthetic content is rapidly eroding. This creates risk for:

• Governments
• Media organizations
• Companies

Many fear they lack secure platforms and adequate training systems to manage this new environment.

Corporate Contradictions

Laurent Félix (Ekimetrics) described the corporate paradox:

• CSR is interpreted inconsistently
• Companies face economic vs ecological contradictions
• Few have genuinely transformed their models

Yet long-term economic performance is increasingly tied to ecological transformation. This gap makes companies vulnerable to misleading narratives.

“Recognizing that economic performance is tied to their ability to transform in an ecological way is fundamental. Few companies have made this shift”, Laurent Félix, CEO Ekimetrics

Brazil as a Disinformation Laboratory

Brazil has become a global testbed. In April 2024, massive campaigns falsely claimed the government was doing nothing about floods. Disinformation is now used as a political weapon. The Rio Summit on Climate Information Integrity raised a central question: what political, legal, and financial tools are actually deployed? So far, the answer remains weak.

Traditional Media Are Not Immune

Contrary to common assumptions:

• People over 60 share more fake news than younger groups
• Young people consume both traditional media and social media
• Traditional media also spread misinformation

Science Feedback found that false information remains more prevalent on social platforms. One key insight: When media actively cover climate disinformation, they increase the amount of verified information. Silence allows falsehoods to dominate.

A Fragile Regulatory Framework

The European Digital Services Act (DSA) aims to structurally constrain platforms. However, the US government is currently pushing to dismantle it, threatening global regulatory coherence. Without binding legal pressure, platforms have no structural incentive to reduce the virality of false content.

What Actually Works

The most effective levers identified so far:

1. Tracing origin sources of narrative
2. AI models specialized in disinformation detection
3. International coalitions (France–Brazil, upcoming biodiversity cooperation)
4. Sanctions for media spreading false information
5. Systematic source citation
6. Stronger coordination between media outlets

Final Assessment

Technology itself is neutral. It amplifies the intent of those who control it. Today, the dominant digital actors are not structurally aligned with the fight against disinformation. Defensive tools exist, but political will and economic alignment lag behind. Younger generations are beginning to understand manipulation mechanisms. But without strong regulation, international coordination, and real sanctions, disinformation will continue to scale faster than the systems designed to stop it. Technology can help. It will not be enough without power shifting.