View this email in your browser

Hello.

Usually this intro would arrive somewhere between a conference stage, an airport lounge and Matt C catching another flight.

Not this week.

The travel schedule finally caught up with the crew, which leaves me, Anika, part of Sønr’s research team, briefly handed the keyboard while the Matts rack up air miles across Scout InsurTech and ITC Europe.

I’m writing this from Durban, mixing research with sunshine and the occasional dolphin spotting on beach walks. Distance has a way of changing how you read the market. It pulls you away from the obsession with novelty and back toward what actually works, focusing on what is accessible, what scales, and what genuinely improves something.

That is a big part of what we think about at Sønr. Not just what is launching, but what actually matters. Who is moving early. Which partnerships signal something bigger. And where the market is quietly shifting before the headlines catch up.

That lens felt especially relevant this week.

Everyone is being told AI can now do research. And to an extent, it can. Ask what is happening and you will get something fast, confident and seemingly comprehensive.

But that is where the problem starts.

AI can answer the question you give it.
It will not tell you whether you are asking the right one.

That gap, between information and understanding, between sounding informed and actually knowing what matters, is where the real work sits. And if I am honest, it is the thing I keep coming back to most when reading the market. Not what launched, but what it means. Not what was announced, but who it actually helps.

That is also why Matt C’s recent conversation with Sønr’s Director of Research, Emily Peters, landed particularly well this week. They unpack the growing gap between information and insight, why domain expertise matters more than ever, and the role of what Em calls “the translation layer” in making AI genuinely useful inside insurance.

A genuinely thoughtful conversation: https://sonr.global/the-translation-layer/

It is also why one of the most interesting insurance launches this month came from Johannesburg.

Naked Insurance introduced what it believes is the world’s first native ChatGPT insurance experience, delivering a binding car insurance quote directly in the conversation. Not an estimate or a redirect, but a fully underwritten quote generated in ChatGPT and connected to its underwriting engine.

That distinction matters. Many conversational journeys still act as entry points, returning an initial quote before handing off to another channel. Here, the underwriting decision and final price are delivered within the interface itself.

More than novelty, it signals something deeper. Teams in resource-constrained, mobile-first markets often move fastest once the infrastructure catches up. Worth watching.

Right, let’s get into it.

From CoPilots To Core Systems

Agentic underwriting continues to evolve. Five stories this week that stand out, and increasingly point in the same direction.

Zurich Insurance completed a 90-day rollout of Cytora’s agentic AI across five countries, reducing manual risk processing time by 80%, with plans to scale to 20+ markets over the next 16 months. This is moving beyond experimentation and into operational infrastructure. A meaningful shift.

That same energy is showing up in how new platforms are being built.

Thrive Capital and Sequoia co-led a $46m Series B into Pace.

Founded by Jamie Cuffe, Pace has processed 250,000+ workflows across Prudential, WTW, Newfront and The Mutual Group.

Not copilots around the workflow. Agents inside the workflow.

What I find interesting about Pace is not just the numbers. It is that the results are showing up in the places that are hardest to move: claims cycles, renewal ingestion, servicing speed. Loving this one.

It is also becoming a priority for core platforms.

BriteCore introduced embedded AI across underwriting, claims and servicing workflows, placing it directly within the core system rather than alongside it. Core platforms are increasingly facing the same question: how deeply to embed AI into day-to-day operations.

Okay, this one caught my attention.

Hippo launched Hannah, an AI-powered service representative, and it is the kind of move that looks simple on the surface but signals something bigger underneath.

Servicing has historically been the part of insurance that AI was supposed to support but never quite owned. Hannah suggests that is starting to change. Customer-facing AI that actually resolves, rather than redirects, is a meaningful step forward.

And that direction is already shaping distribution.

Prudential launched “Just Ask”, giving advisors underwriting and servicing insights in minutes rather than days. Faster access, more responsive servicing, and potentially new ways of supporting the people on the ground doing the work.

Insurance, Where It Didn’t Used To Be

Insurance is starting to show up inside ChatGPT, TikTok and even supermarkets. The distribution conversation is shifting.

Liberty Mutual launched a native ChatGPT experience, enabling customers to explore coverage and receive personalised guidance inside the conversation. Joining Aviva and Naked Insurance, Liberty Mutual is among a growing number of insurers now present within AI assistants. Conversational distribution is starting to feel less differentiated and more expected.

That same shift is playing out in commerce platforms.

ERGO NEXT’s partnership with TikTok Shop is a strong example of embedded insurance done right. It allows sellers to quote and bind cover directly within Seller Center without disrupting their workflow.

It’s a creative but intuitive move, given TikTok’s large and underserved base of creators and sellers. What stands out is the imagination behind it.

The creator economy has grown rapidly, yet insurance has largely overlooked it. ERGO NEXT didn’t, and that kind of thinking is likely to age well.

A similar pattern is emerging in retail.

Allianz UK extended its partnership with Asda to offer home insurance through the supermarket’s platform, bringing it closer to the point of decision rather than relying on customers to seek it out.

Across these examples, a pattern is starting to take shape. The question is shifting from how to distribute insurance digitally to how to show up where decisions are already being made.

A subtle but important shift.

And while distribution is evolving at the front end, the back end is changing just as quickly.

Claims, Reimagined

Great to see our friends at Simfuni expanding into the US and Australia, alongside the launch of an AI-enabled claims platform for life insurers.

Simfuni brings together claims, payments and workflows in a way that works alongside existing systems, not replaces them. As CEO Shaun Quincey puts it, it is “a safe and methodical path to adopting AI tools.” That is still where much of the market sits.

Alongside that, Generali China partnered with Solera to roll out AI-driven EV claims automation across estimation and workflows.

It is starting to feel like a race around data and AI. Teams that understand damage better and get repair vs replace right will win. Still early days, but this feels like a smart, well-timed move.

Fleets, Data And Continuous Underwriting

Static risk models are on borrowed time. A new infrastructure is emerging.

Alliant Insurance’s transportation services arm has launched ASCEND, bringing together commercial fleet leaders across safety, operations, and risk. The focus is not the community. It is the stack. Insurance, telematics, fleet data and analytics are being integrated into one system.

Alliant built both the network and the infrastructure. That combination is harder to replicate and reinforces distribution and defensibility.

Further down the stack, a quieter but more consequential capital story is taking shape.

Infrastructure Becomes The Advantage

In recent weeks, capital has been flowing toward data, intelligence and operational leverage, rather than standalone products.

Tomorrow.io secured a $35m extension to its Series F from investors including Harel Insurance and Pitango, bringing total funding to $210m.

What stands out is what they backing. Not a weather app, but the underlying catastrophe intelligence infrastructure.

Tomorrow.io’s DeepSky satellite network gives carriers more direct access to atmospheric and climate data, which is often slower and more fragmented through traditional sources. This has potential implications for how risk is assessed and priced.

This is close to home. It is the exact conversation we are having with clients right now. Which hazard and climate data partners to choose, and why it matters.

For home and motor insurers, this is not just about data. It shapes how well you price risk. The ones moving early are not just buying better data, they are building an edge that compounds.

A similar shift is starting to appear in health.

dacadoo partnered with Bradesco Seguros’ Vida e Previdência division in Brazil to scale preventative health engagement across the market. Health insurers are gradually moving beyond paying for sickness toward helping people stay well, and Brazil is leaning into that shift.

This is the part of the market I find most exciting. The impact is real. It starts to change what insurance is actually for.

The Rules Taking Shape

The EU AI Act is moving from principle into practice. The European Commission has opened consultations on transparency, deepfake disclosure and governance for high-risk systems. Underwriting, claims and fraud are all in scope.

What our research and client conversations keep surfacing is the tension underneath that. AI adoption inside insurance is accelerating. The governance frameworks meant to sit alongside it are not keeping pace.

That gap is showing up in practical ways. Policy wording was not designed for AI liability. Renewal conversations now include questions that did not exist two years ago. The coverage market is quietly splitting between carriers building affirmative AI frameworks and those still relying on existing language to hold.

The question is not whether regulation is coming. It is who is already building governance around it.

For some, that remains a compliance exercise. For others, it is starting to look like a competitive advantage.

Scout, Barcelona And The Interface Shift

Matt C has been from Bristol to Columbus via Seattle.

He described his route to Scout InsurTech as “deliberately indirect,” designed to maximise time with family before conference season kicked off.

Although, in his own words, probably not a route he would choose again! He landed early Tuesday and went straight into a full day of Scout workshops.

The first time Scout has run as a full-day format, with a clear focus on building and scaling technology within insurance, and open collaboration as the theme throughout.

Scout is run by Chris Luiz, Michael Fiedel and Andrew Daniels, and the quality of personal attention and introductions they deliver at scale stands out.

Wednesday started with a rainy morning run with Peter from Loro before Matt C left mid-session to catch a flight to Barcelona, via Toronto.

He also sat down with Chris Luiz to discuss how Scout is building a year-round insurance innovation ecosystem across the US.

Worth a watch: https://lnkd.in/e4B-GNSr   

Then came Barcelona. Straight off the plane and into panels at ITC Europe. Well, nearly. Airport delays led to a missed the first panel (but thanks Matt F for stepping in🙏). Good energy. Great venue. And plus, by the time you read this, hosted media sessions and ran a roundtable on “The New Innovation Playbook”.

One standout session from ITC was Matt F joined by Arslan Hannani from Aviva and Tim Graham, CEO and Founder of Malcolm, unpacking what conversational distribution looks like in practice. A super interesting case study and one the market should be closely tracking.

Aviva launched a native ChatGPT insurance experience earlier this year. But the more important point is not the interface. It is the architecture underneath it.

Pricing and underwriting exposed through infrastructure. Designed to work across environments, with governance built into the workflow.

Once that layer exists, the same capability can surface anywhere. ChatGPT is simply one example.

Treat conversational distribution as a standalone channel and the case becomes less clear. Treated as infrastructure, the picture changes.

The value builds over time. And the real cost of waiting is not the build, but the learning.

Good to see Matt C joining Generali to judge the AI Liability Challenge Lab, with finalists including Exona Lab, Protostars and Jsonify.

Congrats to Exona Lab and Daniel Reti on the win.

–

That’s me for this week.

A slightly different SøNws. Different geography. Different lens. Hopefully still the same focus on signal over noise.

Writing this from a research perspective, slightly removed from the pace of conferences, announcements and constant updates, changes how you read the market.

The Matts will be back next week with more airport stories, more conference chaos and probably a slightly concerning number of coffees.

And if you haven’t explored Sønr 2.0 yet, it’s worth a look. The companies, partnerships, investments and strategic signals genuinely shaping insurance globally, all in one place. Less noise. Better signal.

As for me, I’m signing off from a very sunny South Africa and heading to the beach before the winter sun disappears. Not a bad way to end the week.

Until next time.

Anika

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.