switzerland Enters the AI Race with Open-Source “Apertus” Model, prioritizing Trust and accessibility
September 2, 2025 – 1:30 PM CEST
Switzerland has officially entered the global artificial intelligence arena with the launch of Apertus LLM, a national language model designed to offer a trustworthy and accessible alternative to dominant players like ChatGPT. While not the first to market – developers acknowledge it’s comparable to Meta’s LLAMA 3 model from 2024, which has since been updated - Apertus distinguishes itself through a deliberate focus on safety, inclusivity, and sovereign control.
[Image of Apertus LLM – Keystone / Ennio Leanza]
The Swiss initiative isn’t attempting to match the massive financial investments of US tech giants. Instead, the team behind Apertus is prioritizing a different set of values. “Our goal is to provide a draft for how a trustworthy, sovereign and inclusive AI model can be developed,” explains Martin Jaggi, professor of machine learning at the École Polytechnique Fédérale de Lausanne (EPFL).
This means consciously avoiding the constant stream of new features often pushed to end-users. The focus remains on building a robust and reliable foundation for researchers and businesses, mitigating risks associated with the rapidly evolving AI landscape.
Addressing Key Concerns: Hallucinations and Bias
The launch of apertus comes at a critical time. The explosive growth of AI as OpenAI’s ChatGPT in 2022 has highlighted significant challenges, including the tendency for large language models to ”hallucinate” – generate false or misleading details – and perpetuate existing societal biases.
The swiss team is acutely aware of these issues and is positioning Apertus as a platform for addressing them. By prioritizing transparency and control, they aim to foster a more responsible and ethical approach to AI development.
Looking Ahead
Switzerland’s entry into the AI race with Apertus signals a growing global desire for alternatives to the concentrated power of a few major tech companies. The project’s emphasis on open-source principles and a commitment to trustworthiness could prove influential as the world navigates the complex challenges and opportunities presented by artificial intelligence.
Keywords: Artificial Intelligence, AI, LLM, Switzerland, Apertus, ChatGPT, Meta, LLAMA 3, EPFL, Machine Learning, Open Source, AI Ethics, AI Safety, Technology, Innovation.