Underwriters plan rapid AI ramp-up to cut admin and sharpen risk assessment
Around 70% of underwriters surveyed by professional services firm Accenture expect their work to be affected by artificial intelligence (AI) within three years. Yet, only 14% are currently using AI.
Around 70% of underwriters surveyed by professional services firm Accenture expect their work to be affected by artificial intelligence (AI) within three years. Yet, only 14% are currently using AI.
“We have been doing research like this for 18 years and we’ve never seen underwriters embrace a new technology like this,” says Matthew Madsen, MD and Global Lead for Insurance Operations Transformation at Accenture. The firm surveyed 430 senior underwriters across life, commercial P&C, and personal P&C insurers in 11 countries. Underwriters plan to use AI to improve data quality, reduce admin, enhance risk assessment, speed up submission intake, and modernize policy platforms.
Alongside the surge in AI demand, underwriting is also being transformed by increasing regulatory complexity, rising customer expectations, looming talent shortages, and rising pressure on insurers to grow, says Madsen.
“Underwriters told us there are three factors that will drive investments in their firms: ease of doing business, underwriting quality and talent.”
Madsen was speaking at a recent online event hosted by the Qorus Insurance Community. The event, which included speakers from AXA, Securian Canada, and Canadian insurtech Foxquilt, examined how AI and automated workflows are changing risk assessment, decision-making, and pricing.
Key takeaways:
- Underwriters expect AI to affect 70% of their work within three years, even though only 14% use it today.
- Underwriters plan to use AI to improve data quality, reduce admin, enhance risk assessment, speed up submission intake, and modernize policy platforms.
- Increasing regulatory complexity, rising customer expectations, looming talent shortages, and rising pressure on insurers to grow are also driving change in underwriting.
- AXA is moving beyond pilots by inserting AI into broker submission ingestion, guideline checks, contract comparison, and selected risk factor detection.
- Securian Canada’s AI guardrails cover privacy, explainability, fairness, and bias controls, delivered through a simple customer experience.
- Foxquilt’s automated workflow aims to rate, quote, and bind a policy within five minutes and its best time is two minutes and 14 seconds.
Half the executives polled at the event confirmed their companies are actively transforming their underwriting while 28% reported their firms are partially modernized. Others have yet to start transforming their underwriting.
Check out the event highlights!
Underwriters burdened by heavy admin demands
Senior underwriters currently spend a third of their time on admin and AI could free them of much of that burden, says Madsen.
“A senior underwriter should be spending 90% or even 100% of their time on applying their technical expertise, understanding the risk and coming up with the best way to help the client, agent and broker and deploying capacity in the best way possible.”
Multinational insurer AXA has already begun incorporating AI capabilities into the underwriting workflows of many of its operating companies.
“Everything around broker submissions and ingestion is now being deployed through AI, so underwriters and operations don’t have to spend time working through lengthy PDFs and emails,” says Yann Bry, Global Head of Business Innovation at AXA.
Bry adds that AXA is also using AI to detect risk factors from policy data, check complex underwriting guidelines, and compare annual broker contracts.
The insurer is using AI to improve its pricing as well. It is employing machine-learning pricing models built by cross-functional teams of actuaries and data scientists and has introduced an automated pricing pipeline to push models into production faster. The company has strengthened its pricing models by pulling in data from additional sources.
Four key lessons from scaling AI
Bry points to four key lessons learned from scaling AI at AXA:
- Business not technology must drive change: Start with what underwriters value most, such as cutting admin, to win early buy-in.
- Improve data quality and accessibility: Without access to quality data AI implementations cannot succeed.
- Prioritize investment by business value: Underwriters should identify business needs and data required so investments target specific outcomes.
- Align underwriting copilots: AI copilots must fit the needs and readiness of each line of business.
Bry adds that insurers need to choose between transforming their underwriting by plugging AI into existing processes or taking the bolder route of completely redesigning their workflows with humans only involved where they add value.
Securian Canada has redesigned its underwriting workflows so that AI quickly processes most cases with underwriters only intervening to address exceptions or complex issues.
“We're moving from a world of static rules and long checklists to something more dynamic where systems that learn support underwriters and help us make better decisions faster,” says Shamir Jamal, Chief Underwriter and VP Business Management Controls at Securian Canada.
Jamal points to the Canadian Automobile Association (CAA), where Securian Canada offers cover to the organization’s seven million members, as an example of the high performance of the insurer’s automated underwriting.
“In our first year with CAA as our client, we achieved 133% of our sales target, with 65% of term life applications, and 99% of the health and dental policies approved instantly. We've seen a 40% increase in overall policy sales and the time taken to issue a term life policy has gone from six to eight weeks to just 20 minutes.”
Securian Canada plans extending its AI automation to improve identification and straight-through processing of low-risk cases. It also intends enhancing its claims processing and customer service with AI supporting predictive models, document reading, medical summaries, and case routing.
Essential guardrails for rolling out AI
Jamal says the insurer’s rollout of AI is guided by firm guardrails.
“For us, that means these key non-negotiables: privacy and explainability, not a black box, clear fairness and bias controls, and a smooth, simple experience for customers.”
Insurtech Foxquilt has harnessed the capabilities of AI and workflow automation to drive rapid growth. The company provides embedded general liability cover to small businesses in the US and Canada. Its technology platform supports an omnichannel distribution network of embedded enterprises, wholesale agents, and direct sales.
“We have consistently doubled our revenue year after year and this has proven our product and our platform,” says Milind Joshi, CTO at Foxquilt.
The insurtech uses AI modeling to analyze loss data when filing and refiling admitted insurance products. Those products need rates and rules approved by local regulators and can take up to three years to change.
“Because these are admitted products, humans really have no leeway in terms of making underwriting decisions. So, we can use AI modeling when we are filing or refiling a product and have done so extensively,” says Joshi.
Foxquilt’s automated workflow aims to rate, quote, and bind a policy within five minutes.
Joshi points out that not all Foxquilt’s processes are suitable for AI automation. The insurtech built an AI chatbot to answer agents’ questions about the types of cover the company provides but withdrew it because the system could not strike a safe balance between rigid and overly creative answers.
“We resorted to including appetite documentation in the broker portal in an easy-to-access way so brokers could read it for themselves instead of us trying to answer their questions.”
Underwriters may expect AI to reshape 70% of their work within three years, but that uptake will only be successful if insurers are clear about which activities should be automated and which need to remain with human specialists. AI is likely to increasingly handle case intake, sorting, and checks, while underwriters will be called on to make and oversee decisions where judgment, fairness, and accountability are critical.
Insurance & Embedded Insurance community
With Qorus memberships, you gain access to exclusive innovation best practices and tailored matchmaking opportunities with executives who share your challenges.