Credit insurance, trade finance and sovereign risk have always shared a common problem: underwriters making decisions on data that arrives late, in inconsistent formats, and from sources that rarely talk to each other. A loan tape from one bank arrives as a spreadsheet, another as a PDF; the same borrower appears under three different names; and a portfolio model that reinsurers need before committing capacity can take a week to build from scratch.
Gallagher Re’s Q3 2026 Global InsurTech Report argues that AI is beginning to change this. Not by replacing underwriting judgment, but by removing the data preparation work that has always consumed so much of it.
The report is part of Gallagher Re’s three-year series on AI in insurance, produced in collaboration with CB Insights. Its thematic focus this quarter is on credit risk protection, trade finance, financial protection and hedge insurance. Those are lines the report describes as historically constrained by “fragmented financial, economic and behavioural data.”
Global insurtech funding fell 44.6% quarter-on-quarter to US$1.36 billion in Q3, according to the report, but the decline was driven almost entirely by a drop in mega-round activity rather than any broad pullback. Deal count was essentially flat at 104 transactions, against 107 in Q2. AI-focused companies attracted 83.2% of all insurtech funding in the quarter, raising US$1.13 billion across 72 deals.
Convergence, a Lloyd’s-based credit non-payment insurer founded in 2023, built its underwriting platform around the premise that credit insurers receive some of the most detailed financial information in the market and do almost nothing with it. Each deal arrives in its own format, the same borrower appears under different names across different banks, and what an underwriting team knows about an obligor typically lives in inboxes and spreadsheets rather than a shared record.
On Convergence’s platform, AI agents process incoming loan tapes and standardise them automatically. A 1,300-row tape that previously took an analyst two to four hours to conform by hand now takes minutes. Portfolio modelling that previously took a week now takes 10 to 20 minutes, according to company-reported figures cited in the report. The platform currently processes more than 500 credit insurance enquiries a month. Convergence grew approximately 140% in 2025, and by the half-year point of 2026 had already surpassed its total for the previous full year.
“Some of the most valuable companies in the world give their product away and monetize what they learn from the people who use it," said Stephen Pike, co-founder and chief executive of Convergence. "Insurers are trusted with data every day and then let it evaporate into inboxes, spreadsheets and PDFs.”
OAK Global, the Lloyd’s specialist underwriter operating through Syndicates 2843 (primary reinsurance) and 1440 (retrocession), is addressing the same problem from the cedant-facing side. Its OCEAN analytics platform connects bordereaux standardisation, obligor matching, loss simulation, reinsurance structuring and portfolio analytics in a single workflow.
The core challenge in credit reinsurance, the report notes, is that apparent diversification can conceal repeated exposure to the same obligor or economic shock. Several cedants may cover bonds backed by the same company without any single player seeing the full picture. OAK Global’s workflow is designed to surface those connections before capacity is committed.
The political risk and trade credit market faces a version of the same problem at a different scale. Ratings and credit spreads are lagging indicators, and for emerging market sovereigns there is often limited liquidity and unreliable data to begin with.
Tellimer, an AI-native data and analytics company, has built a sovereign probability of default model specifically for this gap. Its design combines a logistic regression layer estimating structural solvency risk with a gradient-boosted layer trained to capture idiosyncratic shocks - including commodity swings, data revisions, and abrupt policy shifts - that a structural model alone would miss.
In back-testing, the model identified 26 of the last 28 sovereign defaults, with a 1.8% false-alarm rate, on average more than eight months ahead of the event. For Sri Lanka, Argentina, Ukraine and Lebanon, the model’s signal preceded both bond market moves and ratings changes, according to the report. As of July 2026, global mean sovereign probability of default sat at its highest level in more than two years.
The report’s analysis of Q3 investment data contains one finding that cuts across the technology trend. Re/insurance companies backed 25 technology investments in Q3, down sharply from 51 in the same quarter of 2025. So far in 2026, (re)insurers have made only 59 insurtech investments, against 162 for the full year of 2025.
Incumbent carriers, the report argues, are focusing resources on building their own AI capabilities rather than backing external insurtechs, with the broader environment described as “all hands on deck” for internal AI deployment.
Andrew Johnston, global head of insurtech at Gallagher Re, said the underlying picture in Q3 remained resilient despite the headline funding decline.
“Deal activity was broadly unchanged, early-stage funding recovered, and AI-focused companies continued to attract the overwhelming majority of available capital,” he said. “The more important question is not whether funding is up or down in a particular quarter. It is whether InsurTechs are solving real business problems that insurers, reinsurers and their clients are prepared to pay for.”
Jonathan Allard, Gallagher Re’s global head of credit, surety and political risk, wrote in the report that AI’s best use case in credit underwriting was unlikely to be replacing underwriters. “It will enable the same number of underwriters to achieve higher volumes - and with improved accuracy,” he said. The constraint it removes is data: access to it, consistency of it, and the time previously spent conforming it before judgment could be applied.
