Robots Need Insurance Too

Data centres are getting a new kind of employee: one that doesn't sleep, doesn't unionise, and doesn't need a canteen. It does, however, need something almost nobody in the industry has priced properly yet — insurance built for machines that work alone, at scale, inside critical infrastructure.

A robot for every server rack

Exclaim Robotics, a Zürich-based startup, has just exited stealth with €4.29 million in pre-Seed funding, co-led by Founderful and Playfair. The company builds autonomous mobile robots designed to handle routine repair and maintenance inside AI data centres.

This isn't a novelty product. AI data centres are being built faster than the workforce needed to maintain them, and the physical demands of that maintenance — swapping components, running diagnostics, responding to faults around the clock — are exactly the kind of repetitive, high-precision work robots are good at.

The logic is straightforward:

AI compute demand is growing faster than skilled technician supply.
Downtime in AI infrastructure is extraordinarily expensive.
Robots can work continuously, without fatigue-related errors.

But there's a catch nobody in the pitch deck talks about: what happens when the robot gets it wrong?

The liability nobody has priced yet

A maintenance robot inside a live data centre isn't a low-stakes machine. It's operating near power systems, cooling infrastructure, and hardware worth millions. A single misjudged repair could trigger outages, hardware damage, or worse.

Standard commercial insurance wasn't built for this. Insurers underwriting general liability or tech E&O policies don't have deep actuarial data on autonomous maintenance robots, because the category barely existed until now. That leaves companies like Exclaim Robotics — and their data centre clients — in a grey zone: technically insurable, but not efficiently or affordably so.

This is precisely the gap that captive insurance is designed to fill.

Why the UK's captive insurance proposal matters here

In July 2026, the PRA and FCA proposed a new captive insurance regime aimed at establishing the UK as a serious hub for this fast-growing market.

A captive insurer is, in simple terms, an insurance company owned by the business it insures. Instead of buying cover from a third-party insurer at a price shaped by limited data and cautious underwriting, a company sets up its own regulated insurance entity to cover its own risks.

For high-value, high-uncertainty tech operations — robotics, AI infrastructure, autonomous systems — this model solves a real problem:

Better risk pricing. The company understands its own robots and failure modes better than any external underwriter.
Faster claims and cover for novel risks. No waiting for the broader insurance market to "catch up" with new technology.
Capital efficiency. Retained risk can be managed more flexibly than through third-party premiums.

The proposed UK regime is explicitly framed around growth and competitiveness — a signal that regulators see this as infrastructure for the next wave of industrial technology, not a niche financial product for multinationals.

The connection is not accidental

Exclaim Robotics is Swiss, not British. But the type of company it represents — venture-backed, deploying autonomous hardware into physical, high-value environments — is exactly the profile the UK's captive proposal is trying to attract.

If the UK wants to be where AI infrastructure companies domicile their risk management, it needs a regime that can handle robots doing physical work in data centres, not just theoretical liability on paper. A well-designed captive framework gives founders and operators a way to insure operations that traditional markets are still figuring out how to underwrite.

This is the quiet infrastructure question behind the exciting robotics headline: someone has to pay when things go wrong, and someone has to price that risk sensibly.

What this means in practice

For robotics founders, data centre operators, and insurers watching this space, the practical implications are clear:

Robotics companies entering critical infrastructure should be thinking about risk transfer mechanisms early, not as an afterthought once they scale.
Data centre operators deploying third-party robots need contractual clarity on liability, and captives offer a structured way to allocate it.
The UK regulatory shift could make Britain an attractive base for companies structuring insurance around autonomous hardware risk, even if their operations sit elsewhere.

None of this is glamorous. But infrastructure rarely is — until it's missing.

Key takeaways

Exclaim Robotics raised €4.29 million to build autonomous maintenance robots for AI data centres, addressing a real skills and speed gap in the industry.
Robots operating inside critical infrastructure carry liability risks that traditional insurance isn't well-equipped to price.
The UK's proposed captive insurance regime (PRA/FCA) is designed to fill exactly this kind of gap, giving high-risk tech operators more control over how they insure novel risk.
Robotics and captive insurance are converging trends: as autonomous systems enter physical infrastructure, risk management needs to evolve just as fast as the technology does.