> Quelle: https://souverana.ch/en/insights/ > Sprache: en Insights # AI, explained soberly. Swiss-specific guides to sovereign AI, data protection and architecture, written for decision-makers. ## Provider Data-Protection Check 5 parts 1. [ Part 01 Analysis ### Is ChatGPT GDPR-compliant? What OpenAI's contracts cover A Zurich law firm reviewed OpenAI's contracts: none of them holds up for professional-secrecy work. What that means for business use, and three routes around it the review does not name. 21 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/chatgpt-dsgvo-konform/) 2. [ Part 02 Analysis ### Anthropic Claude and Swiss data protection: what to check Claude works for business data, not for professional secrecy, according to VISCHER. And a promise many organisations rely on has not held for two models since June. 21 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/anthropic-claude-datenschutz/) 3. [ Part 03 Analysis ### Google Gemini & data protection: what SMEs need to know Google leads technically as the only provider, yet stalls on bureaucracy, and a signed addendum does not close the gap while the standard web search stays active. 21 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/google-gemini-datenschutz/) 4. [ Part 04 Analysis ### Using Microsoft Copilot safely: what data protection allows Microsoft closes almost every contractual gap, except for web search. What Swiss companies need to check on Copilot and Azure OpenAI. 21 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/microsoft-copilot-datenschutz/) 5. [ Part 05 Analysis ### Proton Lumo: what privacy-friendly AI looks like Open models, its own servers, encrypted chats: Lumo 2.0 shows what confidential AI looks like. Two catches remain. 9 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/proton-lumo/) ## Law & Regulation 5 parts 1. [ Part 01 Guide ### AI regulation in Switzerland: what already applies Waiting for a Swiss AI law is not worth it. Almost every AI project is already measured against law that has been in force for years. 3 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/ki-regulierung-schweiz/) 2. [ Part 02 Guide ### revFADP and AI: what the law actually requires The fine falls not on the company but on the person who approved the tool, up to CHF 250’000. That makes the approval a matter for management. 4 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/revdsg-ki/) 3. [ Part 03 Guide ### Professional secrecy and AI: what may go to the cloud 'That does not work for us' is the standard answer to AI in law firms and medical practices. Art. 321 of the Swiss Criminal Code says something narrower, and most of daily work falls outside it entirely. 4 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/berufsgeheimnis-cloud/) 4. [ Part 04 Template ### AI governance: one page is enough to start A policy that takes three months to finish protects no one today. Shadow AI grows for exactly as long as the document keeps maturing. 4 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/ki-governance/) 5. [ Part 05 Guide ### EU AI Act Switzerland: role first, then duties Most Swiss companies owe the EU AI Act almost nothing. One single point gets expensive, and whoever slips into it usually notices too late. 4 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/eu-ai-act-schweiz/) ## Sovereign AI 5 parts 1. [ Part 01 Guide ### What sovereign AI actually means for Swiss companies If you sell discretion, you have a problem the moment a client asks where the data goes. Sovereignty is the answer, and it is not a product but a dial. 16 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/was-ist-souveraene-ki/) 2. [ Part 02 Strategy ### AI strategy for SMEs: the questions to answer first Most AI strategies are tool lists. Nine questions that determine the sovereign path for Swiss SMEs, without a ready-made framework. 15 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/ki-strategie-kmu/) 3. [ Part 03 Guide ### Sovereign AI infrastructure: why the model is the easy part The language model is the most replaceable part of your AI. Why embeddings and OCR decide your lock-in, and which Swiss hosters offer every component. 14 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/souveraene-ki-infrastruktur/) 4. [ Part 04 Platform ### Sovereign AI platform: why it binds more than the model The model is easy to swap. The platform holds your chats, files and knowledge, and that is what binds you longest. How to choose platform, host and operations separately, and sovereignly. 13 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/souveraene-ki-plattform/) 5. [ Part 05 Integration ### Sovereign AI integration: when AI starts to act The moment an agent acts instead of merely reading, a mistake becomes an incident. Integration is where your IP is built and where you have to keep control. How that works. 12 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/souveraene-ki-integration/) ## Local Inference 5 parts 1. [ Part 01 Analysis ### Self-hosting an LLM: GPT-5 performance for CHF 8’500 Two RTX 4090s, CHF 8’450, three weeks of measurements. Which tasks a machine of your own can carry, which it cannot, and where the real limit sits. 14 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/llm-selbst-hosten/) 2. [ Part 02 Technical ### VRAM for LLMs: how to size it correctly before you buy How to calculate how much graphics memory you need from parameter count, quantisation, context length and concurrent requests. With a calculator and the measurements from two RTX 4090s. 14 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/vram-llm/) 3. [ Part 03 Technique ### LLM quantisation: how it works and how to use it well Quantisation is the lever that fits a large model onto affordable cards and makes it answer faster. How it works, how to quantise a model yourself, and how to find the level at which quality still holds. 14 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/llm-quantisierung/) 4. [ Part 04 Technical ### Ollama, llama.cpp or vLLM: choosing the right engine Whether Ollama, llama.cpp or vLLM: the choice comes down to two questions. How many people work on the model at the same time, and how much graphics memory do you have relative to the model you want? Three tools, measured on the same hardware. 14 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/ollama-llamacpp-vllm/) 5. [ Part 05 Analysis ### Self-hosting an LLM: the real cost and the break-even Whether a machine of your own pays off against Swiss token prices comes down to one figure: utilisation. The full calculation to check for yourself, from the parts list to the break-even, with measured power instead of estimates. 14 Aug 2026 · Updated 24 Aug 2026 ](/en/insights/llm-selbst-hosten-kosten/) ## More articles 1 part 1. [ Analysis ### What LLMs really cost: the price per token is misleading Comparing price lists is time wasted: token consumption, cache hit rate and quantisation decide what you actually pay. 7 Jul 2026 · Updated 24 Aug 2026 ](/en/insights/llm-kosten/) [Start the AI Sovereignty Check](/check/)