Interview: Relm executive says insurers lack capacity for Bybit-scale crypto hacks

The Bybit hack was a reminder that crypto’s biggest risks are not just technical. They are also financial, legal, and increasingly difficult for insurers to absorb.

While digital asset firms are buying more coverage than before, the global insurance market still lacks the capacity to fully protect against a mega-scale breach.

That gap is now shaping how insurers look at Web3, AI, and other emerging sectors.

Hot wallets, crime risk, AI liability, and even kidnap and ransom cover for crypto founders are becoming part of the same conversation: how do you insure industries where the losses can be severe, fast-moving, and hard to model?

In this interview with Invezz, Christian Davies, Global Head of Distribution and Innovation at Relm Insurance, explains why a Bybit-scale event remains difficult to insure, how liability is shared when AI systems cause harm, and why specialist insurers must understand the full risk stack before offering cover.

Davies also discusses regulation, pricing discipline, and how Relm plans to defend its niche as larger carriers enter the digital asset market.

Christian Davies, Global Head of Distribution and Innovation at Relm Insurance

Excerpts:

Invezz: When an AI system causes financial loss or medical harm, how should liability be allocated among the developer, deployer, and end user? Is insurance a workable solution yet?

Christian Davies: Liability sits across the whole stack, and it depends on who is actually buying the insurance. The developer could create something that doesn’t work properly, in which case the liability sits with them.

The deployer could implement it incorrectly. Often the deployer and end user are the same. For example, a law firm might deploy an AI tool to assist with legal work.

If they use that tool incorrectly or rely on it in the wrong way, they are the ones exposed.

Even where there is a developer issue, you would need to prove how much customization has taken place before pursuing them.

It’s no different to any other technology stack. Insurance is absolutely a workable solution.

We’ve created products that sit alongside traditional liability cover, which may exclude AI, and provide affirmative coverage for AI-related exposures.

Invezz: The Bybit hack showed just how exposed this industry is. What percentage of digital asset firms actually carry meaningful insurance, and why is that number still embarrassingly low?

Christian Davies: A lot of Web3 and digital asset companies do carry insurance, but when you look at events like Bybit, there simply isn’t enough capacity in the global market to cover that level of exposure. Insurers have to manage their own risk.

They are not going to deploy billions in limits where the risk is high and the premiums wouldn’t support it.

For example, cold storage can attract very large limits at relatively low rates because the probability of loss is low. Hot wallet and crime risk is different.

It’s a much more nascent market, and while some see it as expensive, the underlying data supports the pricing.

The most common coverage in the market is still D&O, where companies generally carry meaningful
limits. For crime and asset protection, coverage is more limited.

Invezz: How do fragmented regulatory regimes in the US, EU, and UAE change how you underwrite
crypto risk across jurisdictions?

Christian Davies: We underwrite to the regulatory environment in each jurisdiction. That includes whether there is a clear framework, what the licensing requirements are, and whether there are minimum insurance requirements.

Those factors directly influence how risk is assessed and structured.

Invezz: What does it say about the market that kidnap and ransom cover is now relevant for crypto
founders?

Christian Davies: The market has become more stratified. There are significant realised gains from the digital asset sector over the past few years.

There is also a perception that funds can be moved quickly and easily, which makes individuals in this space a more attractive target. That’s why kidnap and ransom coverage is becoming more relevant.

Invezz: When there is little actuarial history for space payloads or psychedelic therapy clinics, what are you actually pricing, and how do you explain that to a reinsurer?

Christian Davies: We work closely with our clients and build models using their data.

While there may be limited public actuarial history, we have our own experience from supporting these sectors over time.

We also draw on comparable risks from other industries. Where risks are less well understood, pricing reflects that. Unknown risks command higher premiums.

That’s a function of uncertainty rather than anything else.

Invezz: Do you worry about blind spots when you use AI to underwrite AI companies?

Christian Davies: Every type of underwriting has blind spots. If it didn’t, you could quantify everything and there would be no risk.

We don’t rely on AI to underwrite. It’s a tool that supports the process.

Underwriting remains a human-led, judgment-based process. The focus is on identifying and managing as many of those blind spots as possible, which is no different to any other line of insurance.

Invezz: As large carriers become more comfortable with digital assets, how do you see Relm defending its niche?

Christian Davies: Relm operates as a primary market with significant experience in these sectors. Some new entrants are coming in with limited data and are pricing aggressively.

We won’t compete in a race to the bottom where we know where claims originate. We have handled significantly more claims in these sectors than most of the newer markets.

That experience informs how we price and structure risk. Some business will move, and some will return.

Our focus is on staying close to our clients, continuing to develop products, and maintaining the underwriting discipline that comes from that experience.