SafetyWing
builds a country.

The welfare state, rebuilt as a subscription

On any given Tuesday, somewhere on the coast of Portugal, a freelance designer opens a laptop, files a $300 dental claim, and goes back to her cortado. Her insurer does not have a branch office. It has, in fact, no office at all. It is SafetyWing - a Norwegian-founded, Y Combinator-backed insurance company with about 150 employees scattered across roughly 40 countries, none of whom can quite point to a headquarters on a map.

This is not a charming detail. It is the product. SafetyWing exists because borders broke insurance, and somebody had to fix it. The pitch is unusually large: build a global social safety net on the internet - health, income protection, retirement, the whole boring stack - and let anyone on Earth buy in regardless of where they were born or where they happened to wake up.

Borders are a UX problem

The world's safety nets were stitched together in the twentieth century, when work meant a building and citizenship meant your taxes paid for hospitals nearby. Both assumptions have quietly collapsed. By the mid-2010s, somewhere between thirty and forty million people described themselves as "digital nomads" or fully remote workers. None of their governments had any idea what to do with them.

The result was a particular kind of misery that only the laptop class can have. People in their late twenties making good salaries, locked out of national health systems, juggling expired travel insurance, calling brokers in time zones they didn't understand, paying out of pocket for a sprained ankle in Bali. The market had products. None of them fit. Travel insurance assumed a return ticket. Expat insurance assumed a permanent posting. Group plans assumed an employer. Remote workers had none of these.

Three Norwegians and a policy memo

Sondre Rasch had an unusual resume for an insurance founder. He had spent years as a policy advisor inside the Norwegian government, where social protection is less a product and more a national hobby. While later working at the freelancer marketplace Superside, he noticed the people on the platform had none of what he had spent his twenties helping design. He pitched a benefits product. Nobody would build it.

So he built it, with co-founders Sarah Sandnes (CTO) and Hans Kjellby (COO), and entered Y Combinator with a deliberately oversized thesis: the welfare state, as a service, sold to anyone with an internet connection. The first product would be modest. The roadmap would not.

They launched Nomad Insurance in 2018 at a price point that read more like a streaming subscription than an insurance policy - around $45 a month, billed continuously, cancel anytime. The legacy industry assumed they would not survive a single major claim. They did, and then quite a few more.

What you actually buy

The catalog now reads like the founding documents of a small, well-organized nation. Nomad Insurance handles travel medical for the laptop-on-a-beach crowd, in Essential and Complete tiers. Remote Health wraps a single, unified health plan around distributed companies whose employees might live anywhere - one policy, 175+ countries, no spreadsheet of local brokers required. Nomad Citizen, launched in beta in 2025, bundles health, income, and travel coverage into a single membership for entrepreneurs and freelancers.

Then there is Plumia, the company's stranger project - part community, part policy lab - explicitly framed as a draft of an internet country. It has more than 12,000 members, no army, and considerably less bureaucracy than most embassies.

Numbers that aren't quite normal

For an insurance company barely seven years out of YC, the trajectory is not where you'd expect. SafetyWing crossed roughly $24M in annual revenue by the end of 2022, with a team that could fit in a mid-sized restaurant. The customer base spans tens of thousands of individual nomads plus hundreds of distributed-first companies - the kind of buyers who already understand that an "office" is a Notion page.

An ambition larger than a category

It would be tidy to call SafetyWing "the insurance company for digital nomads" and move on. The company is not interested in being tidy. Internally, the mission is described as a global social safety net delivered through software - a Norwegian welfare state, basically, but built in TypeScript and sold by Stripe. That sounds like marketing. It is, mostly, a roadmap.

The argument runs like this: governments built protection because markets did not. Remote work has produced a new class of workers who fall outside any single government's reach. Someone is going to build the protection layer for them. The only question is whether it is a private company with subscriber loyalty or a coalition of slow-moving national systems. SafetyWing has a fairly strong view about which one is likelier to show up first.

The country with no flag and a logo

The interesting thing about SafetyWing in 2026 is not the funding or the headcount. It is that the company has quietly become a stress test for an old idea: that protection should follow people, not addresses. If it works - and the revenue curve suggests it is working - the implications spill out of the insurance category fast. Remote-first companies hire from anywhere because someone now insures from anywhere. Visas and residencies get re-thought because the social-protection argument for them weakens. Plumia stops looking like a side project and starts looking like a draft.

None of this is inevitable. The hardest parts of a safety net - retirement, disability, long-term care - are still ahead. Underwriting at this kind of geographic spread is a genuinely difficult engineering problem. The regulatory map is, as the lawyers say, "rich." But SafetyWing has already done the unfashionable thing of proving that someone will pay $45 a month, in fifteen currencies, for protection without a return ticket. That was supposed to be impossible. It now generates revenue.

Back on the coast of Portugal, the designer's claim clears. She doesn't think about her insurer. She thinks about her cortado. Which is, in the end, the entire point.