As insurers grapple with climate volatility, cost pressures, and digital disruption, artificial intelligence (AI) is emerging not just as a technology enabler, but as a structural force reshaping how Canada’s insurance sector prepares, responds, and evolves. From catastrophe modelling to claims automation and customer service, AI and its newer iterations – machine learning, generative AI (GenAI), and agentic systems – are already influencing the industry's most critical decision points.
According to Puneet Chattree (pictured), insurance industry lead at Accenture Canada, these technologies are no longer speculative and they're delivering value across the value chain.
But the promise of AI in insurance, he cautions, depends on application, realistic expectations, and a focus on customer and business outcomes.
AI’s role in climate resilience can be understood in two distinct but complementary phases: before a catastrophic event and after it happens, Chattree said.
On the front end, AI is being used to improve prediction and pricing. Insurers are leveraging peril data, historical losses, and weather modelling to anticipate how risks like flood and wildfire may evolve over the next 10 to 20 years. This is being further enriched by data from IoT devices and satellite imagery.
Chattree pointed to Accenture’s collaboration with ecosystem partners which focus on combining geospatial intelligence with predictive modelling to support foresight. “They're trying to work together to really build around the prediction component,” he said.
Yet some of the most impactful use cases may not require external data at all. Many underwriting and pricing improvements, Chattree explained, can be made by better utilizing internal claims data – an underused asset in many organizations.
“Just the feed-in of how do we better use the data that we have within our organizations to inform better underwriting and pricing decisions has actually been not as mature in this sector,” he said.
Post-catastrophe, the role of AI shifts to acceleration and efficiency. In recent years, Canadian carriers have begun deploying AI tools to streamline claims handling – from triage and document summarization to straight-through processing and parametric payout models.
Chattree noted that some of Canada’s largest insurers are now linking real-time weather data to insurance products, particularly for repeated perils like hail, and automating responses when index thresholds are met. “We’re starting to see greater adoption of that by some of our top carriers,” he said.
Accenture’s research on AI across industries shows that insurance ranks among the most susceptible to disruption – and also among the best positioned to benefit, he said. In a study conducted with the World Economic Forum Chattree cited, the firm examined over 19,000 tasks and projected that 60% of the insurance value chain could be transformed within five years by a combination of GenAI and traditional AI.
The biggest areas of impact? Claims, servicing (including contact centers), underwriting, and marketing.
For claims teams, AI is reducing manual classification work, allowing for faster triage and routing of complex cases. Tools that can assess photographs, summarize reports, and compare claim inputs without needing multiple adjuster visits are seeing widespread adoption. On the underwriting side, Chattree said AI is streamlining document ingestion and driving workflow automation – steps that improve turnaround time and consistency without eliminating human oversight.
But it’s not all about automation. Many of the most useful tools are augmentative – designed to free up staff to focus on higher-value work. Chattree framed it as a shift in capacity, and not replacement: AI helps skilled professionals make faster, smarter decisions by reducing the clutter of routine tasks.
With any emerging technology, hype can outpace delivery. Chattree acknowledged that AI risks falling into the same expectation spiral as blockchain and the metaverse – technologies once heralded as transformative, but which have seen more limited adoption in insurance.
“There’s always a risk with technology,” he said. “If you look at AI, there is significant advancement... from traditional AI to GenAI to agentic AI. I think this space is absolutely moving really, really fast.”
Unlike previous waves, however, AI is evolving from a base that has been decades in the making. Machine learning, Chattree said, has matured over 30-40 years, and current applications build on that foundation – giving today’s AI use cases more staying power.
“Every one of [our clients] has an AI-based strategy that’s built into their broader strategy,” he said. What began as proof-of-concept is now moving into implementation – especially in areas like contact centers, underwriting, and marketing workflows.
Still, organizations must temper urgency with discipline. “It doesn’t matter what the technology is,” Chattree said. “Stay true to the business value... focus on growth, focus on productivity, focus on cost to serve.”
For all its potential, Chattree believes that AI’s success in insurance will ultimately rest on human factors: trust, training, and change management.
“The one thing that our sector will always have is humans at the center of that,” he said.
Whether the goal is augmentation or automation, understanding the human impact will define whether the investment yields long-term value, he pointed out.
“It’s a trust game,” Chattree said.
