New AI and Big Data Report explores risks and opportunities of adoption and acceptance

The Institute's new research report, AI and Big Data: Implications for the Insurance Industry in Canada, provides an important analysis of the challenges presented by machine learning, artificial intelligence (AI) and big data analytics. The report explores actions required by the industry to ensure responsible adoption over the next 10 years.

AI and big data will make insurance better for Canadians by increasing the industry’s capacity to assess risk at a granular level, identifying low levels of risk and supporting risk management to mitigate losses. At the same time, the industry will need to plan for the needs of those consumers who face reduced availability of coverage and higher costs. Transparency, explainability, and the fair treatment of consumers will be central to managing the coming changes.

The report considers six critical questions and provides eight recommendations for the industry.

Want to read it later? Download the report now.

Download full report

The six questions are:

1. What is big data analytics, and how will it support sound business decisions?

The speed and capability of computers has doubled every two years since the 1970s, enabled by the mass production of integrated circuit chips combined with higher transistor density and lower costs. This incredible growth in processing power has made possible the automation of industrial and manufacturing processes, robotics, the Internet of Things, self-driving cars, cloud computing, as well as artificial intelligence and big data analytics.

The term “big data” was introduced in the 1990s to describe the complex mass storage of data that were too big to be processed by the tools available at the time. Since the 1980s, the world’s capacity for big data has doubled every 40 months. This increase in our ability to store data has been accompanied by a growth in our ability to access and analyze information.

In the not-too-distant past, insurance companies conducted actuarial analysis of recent loss experience to anticipate future claims on a quarterly or annual basis. Insurance companies of the future will utilize a variety of big data tools and approaches, including real-time analytics to understand and meet the needs of their consumers; to visualize trends and to support short and long-term decision-making about claims, pricing and operations.

AI and Big Data banner

2. What is artificial intelligence and how will it support sound business decisions?

Artificial intelligence is the use of computer algorithms to simulate and augment human deduction, reasoning and problem-solving, and is already being used in organizations across the financial sector. Use cases include decision support models to increase the cross-selling of products to customers; trained systems to identify suspicious activity and fraud; and natural language processors that enable chatbots and other communication tools.

AI has the potential to fully harness the power of data to better anticipate and serve the needs of consumers as well as to automate tedious tasks and free up valuable time for our industry’s workforce.

One of the major challenges in the application of artificial intelligence in insurance involves explainability, or the ability to translate the solution an AI system comes up with to humans. The adoption of AI in insurance will require communicating about how decisions are made, to provide transparency for consumers, regulators and other stakeholders.


Black box models in artificial intelligence and machine learning are created directly from data by an algorithm, meaning that even the humans who design them cannot understand how variables are being combined to make predictions.

3. Why are AI and big data analytics expected to transform the industry?

AI and big data are expected to touch every area of insurance and reinsurance, including, brokers and agents, adjusters and appraisers and others. There is potential to offer consumers tailored coverages with accurate pricing, and to apply emerging data science tools for risk management to reduce losses before they happen.

Significant applications exist for big data and artificial intelligence to support underwriting decisions. High-quality datasets mean that companies decide what information is important; some companies take on as much information as possible, while others seek only minimal data points.

Claims management applications include the optimization of process flows, cost estimates, image processing, and loss prediction models, resulting in fewer losses and quicker claims processing and pay outs to consumers. Loss prevention will grow in importance with the expansion of sensors and the Internet of Things to support preventative services for home and car owners.

Opportunities exist in other areas of operations, including investment management, hiring practices and marketing, and will require a thoughtful enterprise-scale transformation that will benefit and improve the performance of the Canadian insurance industry.

4. How can the insurance industry best anticipate and manage emerging risks?

As the industry adopts AI and big data analytics, a focus on insurance consumers and their perspective will be critical to facilitating wide acceptance of new tools, systems and technologies.

Insurance should look to other industries that have undergone rapid change in response to technology and adopt lessons learned on managing risks in areas including:
   1. explainability and transparency;
   2. fairness and bias;
   3. availability; and
   4. privacy and security.

The industry must consider independent supervision of industry practices by regulators as an important factor in building consumer trust.


5. Why should insurers expect increased attention from regulators?

Insurance supervisors in Canada and elsewhere are presently reviewing and seeking to understand the different ways that financial institutions plan to use AI and big data. They are particularly interested in the social acceptability of new practices and how their application will benefit consumers.

Some regulators, including the Organisation for Economic Co-operation and Development (OECD) have specifically endorsed the concept of regulatory sandboxes and innovation hubs, where ideas and approaches can be tested before being more widely applied.

While growing regulator interest and action has the potential to slow the pace of industry innovation, the industry has an opportunity to champion new approaches and demonstrate how they benefit their consumers, to help build confidence in the insurance industry.

6. Why must insurers focus on better outcomes for consumers?

Concerns about trust in insurance are expected to be magnified in the age of AI and big data. More data and better tools can support granular decisions about pricing and coverage, benefiting most consumers – but not all. Undoubtedly new approaches that benefit lower-risk consumers will disadvantage those members in a higher risk.

Consideration of this possible shift reclassifying individual risks must be accompanied by a discussion about fairness; insurance practices must be fair and be seen to be fair.

The near universal use of insurance by car, home and business owners means that the application of AI and big data will be monitored closely and judged by consumers, regulators and others. Opportunities exist for the industry to engage with Canadians to explain changes and verify the fair treatment of all its stakeholders.

Change brings risk – and large change, like that associated with AI and big data analytics, will bring large risk.


The report provides eight recommendations for the insurance industry in Canada to address risks in AI and big data over the next 10 years:
   1. Inform consumers;
   2. Embrace innovation;
   3. Be prepared for uncertain regulation;
   4. Create new insurance programs;
   5. Take the time to do it right (and don’t be scared to fail!);
   6. Respond to changing consumers;
   7. Invest in new technology; and
   8. Accept different views of fairness.

Artificial intelligence and big data analytics will bring benefits for insurance consumers if companies deploy them with sufficient diligence, prudence, and care. The insurance industry should be excited by the opportunity to responsibly bring new ideas and approaches forward to better serve the risk management needs of Canadian consumers.


On April 19, join report author Paul Kovacs, Senior Researcher, Insurance Institute of Canada, and Founder and Executive Director, Institute for Catastrophic Loss Reduction, as he presents the research findings. Paul will be joined by panelists Sven Roehl, EVP msg global solutions Canada & Co-Founder Cookhouse Labs; Sachin Rustagi, Director, Digital (Innovation, Distribution & Labs), Gore Mutual Insurance; and Mark Struck, Vice President, Data Analytics, Wawanesa Insurance.