Content moderation on social media is being defunded and abandoned, prompting several countries to seek alternatives. During the RightsCon conference in Taiwan, experts noted that the withdrawal of US funding—driven by the Trump administration—has weakened global digital rights initiatives. This has also affected US tech companies, which are now investing less in smaller user communities, particularly non-English-speaking ones.

In response, Europe is reconsidering its technological dependence on the US, especially in the field of artificial intelligence (AI). The goal is to develop national alternatives better suited to local contexts. A clear example is content moderation. The large language models (LLMs) used by social platforms, primarily trained in English, fail to detect gender-based violence in countries like India, South Africa, and Brazil. Moreover, these systems tend to amplify errors when applied to other languages, according to AI governance experts.

The underrepresentation of non-Western languages in AI models highlights the need for more localized approaches. This involves creating language-specific models capable of interpreting slang, insults, and hybrid expressions. Initiatives like Shhor AI, which has developed a moderation API focused on Indian vernacular languages, and startups like Lelapa AI, which builds models for African languages, are examples of this shift.

The global landscape is also driving technological sovereignty. During the recent AI summit in Paris, several countries announced sovereign AI projects to gain full control over their AI development. The European Union, for its part, has appointed a commissioner for technological sovereignty and is working on creating its own digital infrastructure, inspired by India’s “India Stack” model.

However, government involvement in AI carries risks. Some experts warn that if governments decide which languages or worldviews to prioritize in training data, it could influence the models’ narratives, affecting diversity and neutrality.

Although it is still too early to measure the impact of this trend, the race for technological autonomy is underway, and Europe is determined to secure a prominent position in AI development.

By: Nestor Castillo, ForAllTechNews Director