The University of Applied Sciences and Arts Northwestern Switzerland (FHNW) conducted a comprehensive test of ten of the most well-known AI chatbots on behalf of SRF’s consumer affairs program Kassensturz. A total of 331 everyday, professional, and creative tasks formed the basis of the comparison. The evaluation followed the Swiss school grading scale (1 to 6), with 6 as the highest score. Copilot received an overall rating of “Good,” highlighting its strengths in the areas where people most commonly use AI today: research, information structuring, decision support and creative inspiration.
Key findings of the FHNW test on Microsoft Copilot at a glance:
- High reliability in practical and creative tasks: Copilot achieved very good to good scores in the categories of health-related questions (5.6), everyday tasks (5.0) and Swiss law (5.1), as well as further “Good” ratings in creative tasks (5.3) and in the knowledge categories related to news, science and technology (4.9 each). This shows that Copilot performs strongly in areas where users require practical day-to-day support.
- Assessment by the testers: In particular, the broad range of features and the deep integration into Microsoft products were highlighted.
“Our goal is to develop AI solutions that deliver real value in everyday life and at work – from research to creative inspiration. Feedback from research and practical use is essential to advancing Copilot safely, responsibly, and transparently, ensuring it continues to meet the evolving needs of users in Switzerland and worldwide,” says Catrin Hinkel, CEO Microsoft Switzerland.
Why these results fit well into Switzerland’s AI landscape
The FHNW comparison was conducted at a time when AI adoption is gaining momentum across Swiss companies. According to the Microsoft Work Trend Index 2025, 52 percent of Swiss organizations are already using AI Agents to automate entire business processes – a figure that exceeds both the global and European averages. In addition, also from the Work Trend Index 2025: 80 percent of Swiss executives considered 2025 as a pivotal year to rethink AI strategically and integrate it into existing work models.
Companies and employees are no longer in an orientation phase, but in an active implementation phase.
Figures from the Microsoft Diffusion Report also show that AI is already widely used in Switzerland, with an active usage rate of 32.4 percent among the working population.
For this reason, the results at hand are particularly valuable. They confirm Copilot’s strengths in categories that matter most in everyday life in Switzerland – both professionally and privately. At the same time, they offer clear guidance on where capabilities can be further strengthened to support more complex tasks more effectively in the future.
Further Information (in German):
Kassensturz Espresso (SRF): AI Language Models Compared
Full Kassensturz Episode: How Good Are ChatGPT, Gemini & Co. Really?