While Brussels builds a dense rulebook around artificial intelligence, Switzerland has deliberately kept its own approach thin. That gap may end up mattering more to where AI companies choose to build than any tax incentive.
Switzerland has built its modern economy on a habit of doing less, deliberately. Where neighbouring economies reach for a new regulatory framework, Switzerland's instinct is usually to ask whether existing law already covers the problem before writing a new one. On artificial intelligence, that habit is becoming a genuine point of difference — and possibly a genuine advantage.
The European Union's AI Act is, by design, one of the most comprehensive attempts anywhere to regulate artificial intelligence by risk category, with compliance obligations that scale according to how a system is used. Whatever its merits, it is unambiguously a large, complex piece of law that AI companies operating across the EU must now build compliance functions around. Switzerland, sitting outside the EU and under no obligation to mirror Brussels, has so far chosen not to.
Switzerland is not arguing that AI needs no rules. It is betting that most of the rules it needs already exist.
This is not a case of Switzerland having no view on AI governance. Swiss data protection law, product liability rules, and financial services regulation already apply to AI systems the same way they apply to any other technology. The Swiss position, articulated by its federal administration, has been to monitor developments, participate in international standard-setting bodies, and avoid pre-emptively legislating for risks that existing sectoral law may already address.
Critics call this complacent. Supporters call it proportionate. Both are, in a narrow sense, correct — but the more interesting question for business is not who is right philosophically. It is where AI companies actually choose to locate research, compliance-sensitive product development, and enterprise sales operations when given a genuine choice.
Anyone who watched the last two decades of technology regulation has seen this pattern before. Ireland's relatively predictable, business-friendly regulatory environment made it the default European base for large US technology companies, well beyond what its tax rate alone would explain. Switzerland is not attempting to be Ireland — its ambitions are narrower, its population smaller, its capital markets less central to Europe. But the underlying mechanism is similar: regulatory predictability and a lighter compliance burden are themselves a form of industrial policy, whether or not a government frames them that way.
A start-up choosing where to build is not just comparing tax rates. It is comparing how many lawyers it will need before it ships a product.
This approach is not without cost. A lighter domestic framework does nothing to exempt Swiss AI companies from the EU AI Act the moment they sell into the EU market, which remains most companies' largest addressable customer base by a wide margin. Switzerland's advantage, in other words, is specifically about where development, research, and early-stage commercialisation happen — not a permanent shield from the compliance obligations that come with scaling into Europe's largest single market.
There is also reputational risk. If a high-profile AI harm originates from a Swiss-based company operating under a lighter domestic framework, the political pressure to import EU-style rules wholesale would be immediate and difficult to resist. Switzerland's bet is a considered one, not a permanent exemption from the debate playing out everywhere else.
For now, though, the calculation looks sound. Switzerland is not trying to out-regulate Brussels or out-spend Washington. It is simply betting that, for a certain kind of AI company deciding where to put its next research lab, a thinner rulebook and a strong university pipeline will outweigh the marginal benefits of being physically closer to any single large market. It is a quiet bet. It may also be the right one.
