View this email in your browser

Hello.

Happy Friday.

I’m loving the 20VC pod at the moment. If you’re not listening to Harry Stebbings, you should be.

This morning’s episode – joined by Jason Lemkin and Rory O’Driscoll – had a line that 100% registered: every year or two, have lunch with the CEO of your competitors.

The context was the news that Elon Musk’s xAI has struck a major compute deal with Anthropic. The enemy of your enemy is your friend n’ all that.

It made me think of when I grabbed coffee with Manlio Carrelli, the former CEO of CB Insights who this week moved on to take the top job at Stensul – an enterprise marketing platform well worth checking out. Manlio is a fantastic guy and I wish him all the luck for this next move.

Sønr, CB Insights & Market Intelligence AI

I’m super interested to see where CB Insights takes its strategy from here and, more broadly, where the future lies for market intelligence firms like Sønr.

Will Claude eat the market?
Will cross-industry generalists have a future?
Will it be all about domain experts?
Do you need hybrid models in the mix?

On the same 20VC pod, they talked about Anthropic being weeks away from launching a legal AI product that goes head-to-head with Harvey and Legora, two businesses we’ve been chatting a lot about on Unfiltered. I also happen to be on stage with Legora at ITC Europe in Barcelona.

The question they chewed through was can a foundation model – Claude, ChatGPT, other, really displace a domain specialist?

The skeptic’s case: When you’re a lawyer billing $2k an hour, a $200-a-month Claude subscription is never going to be the same as a platform trained on a decade of case law, integrated into your workflow. You simply can’t submit briefs with hallucinations in them.

But the counter is equally compelling.

History tells us platform layers expand fast and take more than anyone expects. The companies that survive aren’t building point solutions for individuals – they’re building coordinated enterprise workflows so deeply embedded that no horizontal model can unpick them.

The ones that survive?
Laser-focused on domain, data, and depth.

Which maps almost perfectly onto Sønr 2.0.

Yes, people are using – and will continue to use – foundation models for research and general intelligence. They’re extraordinary. Our research team uses them extensively.

But there is a fundamental difference between general intelligence and domain intelligence.

When you’re an innovation lead at a major carrier, a CVC team evaluating an MGA, or a Head of Underwriting deciding which workbench is worth piloting – you need the right intelligence.

Built on ten years of proprietary data. Curated by people who know this market at every level. And delivered through purpose-built workflows – so you never face a blank prompt, just a briefing that’s already waiting when you need it. Nice.

If you haven’t already: sonr.global/2-0

Right. Let’s get into the news. That might be a record intro spiel.

The State of the Market

Before we get into individual stories, a quick nod to Andrew Johnston and Freddie Scarratt at Gallagher Re, whose Q1 2026 Global InsurTech Report has landed and, as always, is essential reading.

The headline: global insurtech funding hit $1.63bn in Q1 and a record 95.2% of it went to AI-focused companies. A surprise? Bah, not really I guess.

Andrew: “AI and InsurTech are now almost synonymous.”
Freddie: “The silent risk of AI is becoming audible, and ignoring it is no longer an option.”

Two back-to-back quarters at this level mark the strongest funding activity since Q3 2022. The money is back – and it’s going to one place.

Capital Piling Into AI Infrastructure

Let’s get into some of this investment.

Five big rounds. Across five days.
Each one hitting a different layer of the insurance stack.

Reserv led the way with a $125m Series C.

Founded in 2022, Reserv is now the largest AI-native TPA in P&C – $100m ARR, ~200 insurer clients, 500,000 complex claims processed per year.

And Blitzy raised $200m at a $1.4bn valuation – and the story here is who backed it: Liberty Mutual Strategic Ventures and Erie Strategic Ventures.

Blitzy builds dynamic knowledge graphs of enterprise codebases and deploys thousands of AI agents in parallel to autonomously execute entire software modernisation projects.

The problem it’s solving for insurance?
Decades of legacy COBOL-based policy administration that traditional IT  consulting takes years and hundreds of millions to fix. And not always too  brilliantly.

Blitzy supposedly compresses that to months.

The New Generation of AI Carriers

The carrier and distribution layer got just as much attention.

Corgi hit unicorn status this week with a $160m Series B led, valuing the business at $1.3bn just four months after a Series A valued it at $630m. You’d be happy with that return if you were part of their Series A round.

Founded by Nico Laqua and Emily Yuan (former OpenAI product manager), Corgi is building the first AI-native, full-stack insurance carrier for startups – property, liability, D&O – and expanding into trucking, payroll and small business.

This week they also launched a standalone AI insurance coverage product, integrated directly into existing Tech E&O policies, protecting businesses from the financial and legal exposure of AI going wrong.

One to watch VERY closely.

In other news, Counterpart Insurance closed a $50m Series C.

It’s a LA-based, AI-native MGA and specialty carrier, pioneering what it calls “Agentic Insurance” – a continuously learning platform across underwriting, broker support and claims.

And rounding out the week, Outmarket AI raised a $17m Series A – AI software for insurance brokers and agencies that connects directly with agency management systems to turn brokerage data into a centralised intelligence platform.

AI Embedded Where the Work Happens

Two announcements this week that quietly matter more than the headline rounds.

Verisk and Anthropic embedded Verisk’s insurance analytics directly inside Claude – Verisk Underwriting Intelligence (ISO loss cost trends, filing signals, conversationally) and Verisk XactRestore (property repair pricing for claims and restoration).

CEO Lee Shavel: “Trust is the foundation of insurance, and that doesn’t change as new technologies emerge.”

The logic is clean – Verisk brings the governed, regulatory-grade domain data; Anthropic brings the interface.

Estimated savings: hundreds of hours per carrier per year on underwriting; 30 minutes to 2 hours per estimate on the restoration side.

On the claims side, insured.io launched Claims AI – a conversational agent handling FNOL end-to-end across voice and chat, with real-time policy search and direct submission into core systems.

I’m a big believer in what voice tech is going to do to customer interactions in insurance, and this is a meaningful step in that direction.

And sitting above both: Allianz and Anthropic’s global partnership – announced in January – continues to build.

Claude available to all Allianz employees, Claude Code for all developers, custom agents for workflow automation, full interaction  logging for regulatory transparency.

The MGA Moment

Antares Global launched Scorpius MGA this week – a new platform to consolidate its MGA relationships under a single structure, led by CEO Ian Morris.

It combines capacity access, shared services and governance in one place, and crucially isn’t tied to a single capacity provider, meaning business outside the Antares appetite can still be placed with third parties.

MGAs are a space we’re tracking very closely at Sønr – demand from clients wanting deeper intelligence on the MGA landscape has grown significantly, from all around the world, and platforms like Scorpius are exactly the kind of structural move that signals where the market is heading.

If you’d like to understand what we’re seeing across the MGA ecosystem, get in touch.

On the Road

It was good to have made it to #GRS2026 earlier in the week.

The Global Risk Summit was at the Guildhall Art Gallery (sitting above London’s Roman amphitheatre built in AD 70) – a day going deep on risk, geopolitics, marine, NatCat, and the uncomfortable truth that incumbents may be falling behind what their clients actually need.

It’s also been a good week catching up with a couple of friends – Óscar Garza, President of InsurTech México, and Finn Magee, CEO of Insurtech Canada – all part of Sønr‘s world domination plans for later this year.

Looking forward, it’s the Scout InsurTech Conference in Columbus, Ohio on May 26-27. Genuinely one of the best formats on the circuit – I loved it last year and cannot recommend it highly enough.

Then straight on to ITC Europe in Barcelona which I am VERY excited about. You’ll see a tonne of Sønr representation and brand activation out there – from a new European Beyond Boundaries Report (launching Tuesday), a media set up for some on-site interviews, a lesser-spotted Sønr booth (you’ve got to try these things right?), a few panel sessions, and as of this morning, I’m going to join the Generali folk as a judge for their ITC Challenge Lab. This time round with a focus on AI Liability – definitely one to come along to.

Right, I’ve got to wrap up.

I have a[nother] flight to Barcelona I have to get to. This one’s for a Hyrox – a test run ahead of the World Champs in June.

Wish me luck 😬

For now, big love. Have a great weekend.

Matt

Make smarter decisions with AI-driven market intelligence.


Sønr helps (re)insurers around the world explore market trends, track competitor activity, and scout and assess best-in-class tech companies.