Intact Financial
AI models for pricing, risk selection and underwriting
Intact built about 600 in-house AI models and says they now bring over C$200 million a year
Problem
The Toronto-based property and casualty insurer competes on how precisely it prices risk. Pricing, risk selection and commercial underwriting depend on large volumes of data and on underwriters filling in quote information by hand.
Approach
Over about a decade Intact built an in-house team of more than 600 data scientists, actuaries and engineers and deployed close to 600 AI and machine learning models, mainly for pricing and risk selection. Machine learning now supports more than 90% of its Canadian personal lines pricing and 60% of commercial lines pricing. It later added a generative AI tool, Underwriting Advisor, in Canadian commercial and specialty lines.
Results
Takeaway
An in-house team built many predictive models for core pricing over years, accounting for most of the value. Management's estimated recurring benefit rose from more than C$150 million in late 2025 to more than C$200 million in the 2025 annual report, without a disclosed calculation method.
Categories
- Size
- Large enterprise
- Industry
- Finance and insurance
- Business function
- Finance, risk and legalInternal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLGenerative AI
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Coverager reports, citing Intact's 2025 annual report, that more than 600 specialists supported close to 600 AI and machine learning models, with recurring annual benefits over $200 million. Machine learning supported more than 90% of Canadian personal lines pricing and 60% of commercial lines pricing. It adds that the generative AI tool Underwriting Advisor raised quote processing by 24% and pre-filled 56% of underwriting data fields.
The Canadian business outlet reports from Intact's third-quarter 2025 earnings call. CEO Charles Brindamour said AI used in pricing and risk analysis was tied to roughly $150 million a year, and that an in-house AI underwriting tool let one specialty line covering unusual real estate investments quote about 20% more. New models had also been rolled out in personal property and commercial fleet insurance.
Full details are in the original sources.