
Lukas Huber
Founder & AI Strategist
OpenAI and ChatGPT under scrutiny: Swiss SMEs must now understand geopolitical risks and data protection concerns when using AI tools.
The accusation is serious: OpenAI, the company behind ChatGPT, is allegedly becoming a pawn in US politics, even an instrument of the Ministry of Defence. Headlines like those recently seen in a Heise analysis can be unsettling. While this might sound distant to Swiss SMEs, the implications are direct and relevant. Geopolitical tensions and data protection concerns can quickly lead to a boycott or at least make the use of standard AI solutions seem too risky.
Especially in Switzerland, where only 12% of companies use machine learning, the question arises: Do we have to forgo the benefits of Artificial Intelligence if the major players suddenly come under fire? The answer is a clear no. Those who remain inactive now and avoid AI altogether risk falling behind. It's not about boycotting the technology itself, but about making conscious and secure choices of tools.
The challenge for Swiss SMEs lies in understanding the potential risks – particularly concerning the strict Swiss Data Protection Act (DSG) and the security of sensitive business data – while simultaneously realising the enormous efficiency gains of AI. A boycott of ChatGPT might be a reaction, but a strategic alternative is the real solution. We need to understand what options exist and how we can integrate them into our business processes safely and in compliance with regulations.
📊 Facts at a Glance:
- Only 12% of Swiss companies use machine learning. (Source: FH HWZ, 2026)
What are the data protection-compliant alternatives to ChatGPT for Swiss SMEs?
There are several powerful alternatives that are specifically tailored to the data protection and sovereignty requirements in Switzerland, or at least offer a significantly better foundation for them than the predominantly US-centric models. The choice heavily depends on specific use cases, budget, and internal IT infrastructure. A blanket switch is rarely the best strategy; rather, a differentiated approach is necessary.
First, it's important to distinguish between different types of AI models. There are proprietary models from large providers, open-source models, and those that allow for on-premise installation or operation in a Swiss cloud. The biggest challenge for Swiss SMEs when using standard AI solutions like ChatGPT often lies in the uncertainty regarding data location and data usage. Sensitive business secrets or personal data must not be transferred to third countries without control or used for training generic models without transparency and contractual agreements.
A promising proprietary alternative, for example, is Anthropic's Claude. This model has seen strong growth among paying consumers in recent years and explicitly positions itself as a security- and ethics-oriented option. While it still lags behind ChatGPT in overall preference, it often offers comparable performance for many use cases. The crucial advantage here can lie in contractual assurances and the ability to better control data flow, depending on the enterprise solutions and hosting options offered. However, it is essential to carefully review the terms and conditions and data protection agreements to ensure they comply with the DSG requirements.
Furthermore, open-source models are gaining increasing importance. Models like Llama from Meta, Mistral AI, or Falcon offer the flexibility to be operated on your own servers or in a trusted Swiss cloud environment. This gives SMEs full control over their data. The downside here is often the higher initial effort for implementation, maintenance, and the necessary expertise. But precisely therein lies an opportunity: Customisation to specific company data can lead to tailor-made solutions that provide a significant competitive advantage.
For SMEs looking to venture into the AI world but have concerns about established US providers, there are therefore various paths available. It is important to evaluate not only the model's performance but also the underlying infrastructure, contractual terms, and the possibility of data sovereignty. A Swiss hosting partner offering specialised AI solutions can bridge this gap and reduce complexity.
| Feature | ChatGPT (OpenAI) | Claude (Anthropic) | Open-Source Models (e.g., Llama 3) |
|---|---|---|---|
| Primary Focus | Broad applicability, user-friendliness | Security, ethics, conversational ability | Flexibility, customisation, data sovereignty |
| Data Location | Primarily USA, precise control often unclear | USA, with stronger assurances for enterprise clients | Selectable (e.g., Swiss hosting) |
| Data Protection (DSG Compliance) | Potentially challenging without specific enterprise contracts | Better starting point due to focus on security, but requires careful review | High potential with correct setup and hosting |
| Customisation | Possible via APIs and fine-tuning, but limited | Similar to ChatGPT, with a focus on secure customisation | Very high, full control over model and data |
| Implementation Effort | Low (SaaS model) | Low (SaaS model) | High (Infrastructure, expertise) |
| Cost Model | Subscription, pay-per-use | Subscription, pay-per-use | Infrastructure costs, development effort (no license fees) |
💡 Recommendation: Check Data Residency
Before choosing an AI solution, request clear written assurances from providers regarding data location and usage. A provider who cannot make a binding statement about hosting in Switzerland or the EU should be approached with caution, especially if you are working with sensitive company or customer data.
How can Swiss SMEs minimise the risks of a ChatGPT boycott while still benefiting from AI?
Through strategic planning, diversification of AI tools, and a consistent focus on internal processes and data sovereignty. A blind boycott without an alternative strategy is counterproductive. Instead, SMEs should develop a well-founded AI strategy tailored to their specific needs and Swiss legal framework. This begins with a systematic analysis of internal processes.
My experience shows: Many SMEs see the benefits of AI but don't know where to start. A structured approach helps here. We must first identify strengths and weaknesses within the company (internal analysis) and then the opportunities and threats in the market (external analysis). A SWOT analysis, combined with a T.O.W.S. framework, is the ideal tool here. After this analysis comes the actual planning. It's about finding concrete use cases where AI creates real added value – whether through automation of repetitive tasks, improvement of customer service, or optimisation of decision-making processes. This systematic analysis to identify AI use cases is crucial for focusing on the internal company environment and not being distracted by external hype topics.
A key pillar in this is building internal AI readiness. This includes not only the technological infrastructure but also the understanding and skills of employees. Swiss SMEs identify a lack of AI specialists and a lack of understanding as central challenges. This is an internal problem that can be solved through targeted training and building internal knowledge. It's not necessary to hire an external AI expert for every task; often, it's enough to empower existing employees to use AI tools effectively and safely.
Diversification is another important point. Instead of relying on a single AI provider, SMEs should pursue a modular strategy. This means using different tools for different tasks. For sensitive data, a locally hosted open-source model might be the best choice, while for less critical, general tasks, an established cloud provider could also be considered, provided the contractual terms are right. This flexibility minimises the risk of becoming incapacitated in the event of an outage or a provider boycott.
Implementation should be done gradually, starting with pilot projects. This allows for gaining experience and optimising processes before rolling out AI solutions company-wide. This iterative approach allows for early identification of risks and making adjustments. A core aspect of this is the complete documentation of all AI deployments, especially concerning the DSG. Transparency about the models used, the data processed, and the protective measures is not only a legal necessity but also builds trust with customers and employees.
🔑 Tip: Build Internal Competencies
Invest in the further training of your employees in the field of AI. A solid internal understanding of the technology and its safe application is invaluable and reduces dependence on external specialists. Start with small workshops and application examples that directly simplify daily work.
What are the concrete impacts of a ChatGPT boycott on the competitiveness of Swiss SMEs?
A blanket boycott without strategic alternatives leads to a significant competitive disadvantage and innovation stagnation. The reality is that AI tools – whether ChatGPT or an alternative – enable immense productivity boosts. Those who do not leverage these advantages will inevitably fall behind. The automation of routine tasks, acceleration of idea generation, personalisation of customer interactions – these are all areas where AI is already delivering tangible results today.
Let's consider the customer side: Kantar forecasts that by 2026, one-third of consumers will buy directly through generative AI platforms. This shows how profoundly AI is already changing customer expectations and behaviour. If Swiss SMEs do not use these channels and technologies, they are leaving the field to competitors who are exploiting these opportunities. It's not just about internal efficiency, but also about direct customer engagement and sales.
The shortage of AI specialists is already a known challenge in Switzerland. A boycott of widely used tools like ChatGPT only exacerbates this problem. It would mean that companies not only avoid the technology itself but also lose access to a broad ecosystem of knowledge, use cases, and developments. This slows down innovation capacity and makes it difficult to develop a well-founded AI strategy in the first place.
Burying one's head in the sand regarding AI is not an option. The world keeps turning, and the potential of AI is too great to ignore. Instead of fearing the risks, Swiss SMEs must learn to manage them. This means actively engaging with the requirements of the DSG and implementing solutions that meet these requirements. It's a matter of responsibility, not refusal. Those who do not optimise their processes with AI will, in the foreseeable future, have higher operating costs and act more slowly than competitors who use AI intelligently.
The greatest danger of an ill-considered boycott is the loss of agility. In a rapidly evolving market, the ability to quickly adapt and utilise new technologies is crucial. A boycott of ChatGPT may provide a sense of security in the short term, but in the long run, it isolates Swiss SMEs from global innovation streams. The message must be: Use AI, but do so wisely, securely, and in line with Swiss values and laws.
⚠️ Warning: Inaction is the Greatest Danger
Fear of data risks or geopolitical entanglements must not lead to a general rejection of AI. The competition is not sleeping. Those who do not act now and explore secure AI alternatives will feel a significant competitive disadvantage in a few years. Losses in efficiency and innovation stagnation are the direct consequences.
🎯 Practical Example: The Swiss Machine Manufacturer
A medium-sized Swiss machine manufacturer faced the challenge of processing customer service inquiries more efficiently. Instead of relying on ChatGPT, the company opted for an open-source LLM operated on its own servers in a Swiss data centre. By fine-tuning with company-specific data, a chatbot was developed that automates 70% of standard inquiries. Employees were relieved, response times were reduced by 12 hours per week, and sensitive customer data remained 100% in Switzerland. The company benefited from AI without compromising on data protection.
Conclusion: Strategy Instead of Boycott
The debate surrounding a ChatGPT boycott clearly shows that the adoption of Artificial Intelligence is far more than just a technical issue. For Swiss SMEs, it's about balancing innovation and security, global connectivity and local data sovereignty. A blanket boycott is not a sustainable solution; rather, a well-founded strategy is necessary that leverages the opportunities of AI while meeting Switzerland's specific requirements for data protection and ethics. There are ways to benefit from AI without compromising on security and compliance.
Three crucial insights for your SME:
- ✅ Explore Alternatives: Powerful, data protection-compliant AI solutions exist that meet the requirements of the DSG. A thorough analysis of options is essential.
- ✅ Manage Risks, Don't Avoid Them: Develop a strategic AI roadmap based on internal analysis and prioritising the development of expertise. Diversify your AI tools.
- ✅ Secure Competitiveness: Inaction is the greatest danger. Those who ignore the potential of AI now will lose out. Use AI wisely and securely to optimise your processes and strengthen your market position.
Would you like to learn how your SME can leverage the benefits of AI securely and in compliance with data protection regulations? Contact us for a no-obligation initial consultation.
Related Articles
Newsletter
Receive our weekly briefing on Swiss AI & Deep Tech.