Episode #181
July 10, 2026
AI Liability, Insurtech’s Next Act, and Why Risk Never Scales Like SaaS
with Freddie Scarratt from Gallagher Re
SPONSORED BY
Freddie Scarratt is Global Deputy Head of InsurTech at Gallagher Re, where he co-chairs the firm’s AI risk initiative. He began his career on the Lloyd’s graduate scheme, with secondments to Beazley in claims and Liberty Specialty Markets in financial-lines reinsurance, before joining the startup reinsurance broker Capsicum Re, later acquired by Gallagher to become Gallagher Re. He now advises insurers and reinsurers on the emerging risks created by artificial intelligence. Connect with Freddie on LinkedIn.
All Episodes Available On
Freddie Scarratt, Global Deputy Head of InsurTech at Gallagher Re, joins James Benham and Rob Galbraith to unpack why 95% of all insurtech funding in Q1 2026 flowed into AI-focused companies — and what that means for the birth of an entirely new risk class: AI liability. From the Air Canada chatbot case that made headlines, to the hard lesson the industry learned when cheap venture capital tried to scale insurtechs like SaaS businesses, Freddie brings a reinsurance broker’s view of where AI liability, cyber, and professional indemnity are converging into one digital risk category — and why insurance, unlike every other industry, can never fully cap its downside.
AI Liability Is No Longer Theoretical
The insurtech market, Freddie argues, has shifted from an experimental stage to execution — and AI sits at the center of both the innovation and the capital allocation. Gallagher Re’s own insurtech report found that 95% of all insurtech funding in Q1 2026 went into AI-focused companies, effectively making the terms “AI” and “insurtech” synonymous for the moment.
That concentration of capital has produced an entirely new product category: AI liability, sitting inside a broader “digital risk” class that didn’t exist as recently as 2019. The clearest proof it’s real, not hypothetical, is the Air Canada case: a customer used the airline’s chatbot to ask about a bereavement fare, the bot hallucinated a policy that didn’t exist, and a tribunal held Air Canada liable for the advice — even though Air Canada didn’t build the chatbot.
“If you’re a deployer of AI and you offer advice based on its output, you could be found liable for that advice.”
Freddie expects standalone AI liability policies, buyback endorsements on existing liability or cyber policies, and increasing reinsurance interest to follow — Munich Re has already begun offering standalone coverage, and Freddie has tracked over 2,000 state filings for AI-related exclusions among major U.S. insurers.
Why the Insurtech Capital Wave Crashed — and What Came Back
Freddie’s team tracks insurtech investment quarter over quarter, and the pattern from 2019–2021 is unmistakable: a flood of non-specialist venture and private equity money entered insurtech the same way it entered every other fintech category, without understanding how risk-originating businesses actually work.
“A thousand bad risks is, if anything, worse than a hundred bad risks.”
Risk doesn’t behave like a commodity that gets cheaper with volume. Writing more policies faster doesn’t dilute bad risk — it compounds it. When that capital realized loss ratios don’t self-correct just by scaling, much of it left. What’s replaced it, Freddie says, is smarter: specialist capital that understands the business, plus insurers and reinsurers posting their two highest-ever years of direct insurtech investment.
From Disruption to Partnership
The early insurtech narrative — 2014 to 2016 — was built around disruption: insurtechs positioning themselves to replace brokers, carriers, and legacy processes outright. Freddie says that framing has aged out.
“It’s not about disruption anymore. It’s about deeply integrated partnerships — how can we help each other get better.”
The model now: carriers and reinsurers provide risk capital, underwriting expertise, and regulatory infrastructure. Insurtechs provide the technology that transforms how that capital and expertise get deployed. Delegated underwriting through MGAs is one visible result — portfolio underwriting teams inside London syndicates increasingly exist specifically to write risk into MGAs run by domain experts.
Why Insurance Moves Slower — And Why That’s Not a Flaw
Freddie pushes back directly on the “insurance is slow to innovate” critique often lobbed at the industry from outside.
“Every other industry has a max downside. In insurance, your downside is unlimited — so you’re always going to be slightly slower to adopt.”
A software company that makes a bad bet loses, at most, what it invested. An insurer that mispreces a risk can face losses with no ceiling — a loss ratio of 10,000% is structurally possible in a way a capped-downside business never experiences. That asymmetry, Freddie argues, is the real explanation for why insurance is measured rather than fast — not a lack of appetite for innovation.
Key Takeaways
- AI liability is becoming its own insurance class. The Air Canada case proved it isn’t theoretical, and the category is expected to converge with cyber and professional indemnity into a broader digital risk class.
- Scale doesn’t make risk cheaper. The 2019–2021 capital wave assumed risk-originating businesses could scale like SaaS. They can’t — more bad risk is just more losses.
- Insurtech has moved from disruption to partnership. Carriers bring capital and regulatory infrastructure; insurtechs bring the technology. Both sides benefit.
- Insurance’s unlimited downside explains its caution. Unlike every other industry, insurance can’t cap its losses — which makes measured adoption a rational choice, not a shortcoming.
Connect with Freddie Scarratt
LinkedIn: https://www.linkedin.com/in/freddiescarratt/
Company: Gallagher Re
Enjoyed this conversation? Check out our previous episode on innovation at scale with Danilo Raponi, Group Head of Innovation at Generali.