How catastrophe losses and AI are shaping Canada's insurance landscape

Brokers, beware of algorithmic bias

How catastrophe losses and AI are shaping Canada's insurance landscape

Catastrophe & Flood

By Nicole Panteloucos

With the new year fast approaching, two major trends will continue to shape the Canadian insurance landscape: the escalating environmental crisis and the rise of artificial intelligence. Devastating events like the wildfires in Jasper, the Calgary hailstorm, and severe flooding in Toronto highlight the need for insurers to adapt to increasingly unpredictable weather patterns and natural disasters.

At the same time, AI presents new opportunities for insurers and brokers to boost productivity, leverage data, and mitigate risks (including those related to weather). However, concerns about algorithmic bias in decision-making and risk assessment persist. Robert Paxton (pictured), global head of strategy and business development and managing director, Canada, at Charles Taylor, alongside leaders from Insurance Business's Power Panels, shared insights on how the insurance industry can navigate critical environmental and technological shifts.

Navigating Canada’s catastrophic environmental losses

Canadian insurers faced record-breaking natural catastrophe losses in 2024. The Insurance Bureau of Canada reported major catastrophic events over the summer, resulting in significant insured losses:

  • $940 million from flooding in Toronto and southern Ontario
  • $880 million from the Jasper wildfire
  • $2.8 billion from the Calgary hailstorm
  • $2.5 billion from flooding in regions of Quebec

Looking ahead, Paxton expects catastrophic risks to remain a top priority for brokers and insurers. "In Canada, we’ve seen the largest natural catastrophe year on record, with over $7 billion in damages so far," he noted. "Environmental risk is front and centre and will continue to impact how we assess and price coverage."

In light of this, brokers must take a more proactive role in helping clients strengthen their risk mitigation strategies. This shift is essential not only in reducing the burden on property and casualty insurers but also in ensuring that insurance is seen as a tool for recovery, not the first line of defence.

During a recent Insurance Business Power Panel, discussing the aftermath of the Jasper wildfires, Kevin Lea, president of Fuse Insurance, pointed out that the responsibility for building nat cat resilient structures should not fall solely on insurers, as this would unfairly raise premiums for those who are already proactive in risk management.

“The cost of building back better needs to come from the homeowners who are benefiting from it and potentially from governments who are looking to risk-proof their populations,” he shared.

In addition to advising clients to reinforce their homes with weather-resistant materials like impact-resistant shingles and reinforced siding, Amanda Modica, director and head of environmental and professional broking at WTW, highlighted the crucial role data plays in managing client risk—an area where brokers can add considerable value.

“Brokers and insurers have access to tools and services designed to help their clients,” Modica explained. “There are platforms that provide data analytics not only on historical loss experiences but also live event tracking for severe weather. Incorporating these tools into emergency response plans can really help you get ahead of things.”

Addressing challenges of AI and algorithmic bias in insurance

“I don’t think anyone fully understands how AI will reshape insurance models,” said Paxton, acknowledging the rapid evolution of AI capabilities. While AI is already helping insurers make more informed risk decisions and uncover new business opportunities, Paxton emphasized the need for caution due to growing concerns over its potential to perpetuate misinformation and bias.

In another Insurance Business Power Panel, Neal Jardine, global director, cyber risk intelligence and claims at BOXX Insurance, pointed out, “AI can create information and learn things beyond what we input. There is a risk that AI can perpetuate bias, as the language models it’s trained on may contain biases, which can then be reflected in its outputs.”

He’s right. Insurers like State Farm have faced lawsuits over racial discrimination in homeowners' claim settlements. Similarly, ProPublica reports have revealed significant disparities in car insurance premiums in Illinois, with drivers in minority neighborhoods paying up to 30% more than those in predominantly white, similarly risky areas.

Tips for brokers on ethical AI usage

  • Audit AI for fairness and bias: Regularly evaluate AI tools to detect and eliminate biased outcomes, ensuring fair treatment for all clients.
  • Protect client data: Avoid entering sensitive client data into public AI platforms, such as ChatGPT. Instead, always use secure, internal systems to protect data and ensure compliance with privacy regulations.
  • Maintain transparency in AI use: Clearly inform clients how their personal data is used in AI-driven processes like pricing, claims, and risk analysis, including the benefits and limitations.
  • Leverage AI to enhance human expertise: Use AI to automate labour-intensive tasks and generate insights, while relying on human judgment for complex decisions to ensure accuracy and enhance the client experience.

Highlighting the crucial role of human involvement in AI-driven processes, Paxton added: "Ultimately, humans should focus on areas where their expertise can add the most value. AI can accelerate processes and handle tasks that aren't as engaging, allowing professionals to spend more time on what truly matters.”

Want to hear from industry experts on the latest trends? Click here to access the full list of Insurance Business TV Power Panels.

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!