Canadian Underwriter

Category: Tech

  • How collision claims are trending for electric vehicle repairs

    How collision claims are trending for electric vehicle repairs

    Technician using a tablet to analyze and diagnose and electric vehicle engine

    Despite slower battery electric vehicle (BEV) sales in North America last year, claims for repairable electric vehicles in Canada and the U.S. keep rising, according to a new report from Mitchell, an Enlyte Company.

    “Even as BEV adoption slowed in North America last year following the end of government tax incentives [in Canada and the U.S.], the auto insurance and collision repair industries still saw claims volume rise since more of these automobiles are on the road than ever before,” says Ryan Mandell, Mitchell’s vice president of strategy and market intelligence, in a press release.

    In 2025, BEV purchases in the U.S. declined approximately 2% from 2024, according to Mitchell’s 2025 year in review Plugged-In: EV Collision Insights report, released Thursday. It’s not clear if there was also a drop in Canadian BEV purchases.

    Still, repairable EV claims also increased in both countries, particularly in Canada. Claims frequency for collision-damaged BEVs in Canada last year represented 4.77% of all repairable auto claims, 4.44% for mild hybrid electric vehicles (MHEVs) and 1.55% for plug-in hybrid electric vehicles (PHEVs). This was a dramatic increase of 24% for BEVs, 29% for MHEVs and 26% for PHEVs in the last 12 months, the report says.

    In the U.S., BEVs represented 3.07% of all repairable auto claims, up 14.1% from 2024. MHEVs came in at 4.82% and PHEVs at 0.89%, a 20% and 6% year-over-year increase, respectively.

    “Due to their dense electrical architectures, software-driven systems and interconnected, sensor-heavy designs, these vehicles require additional diagnostic and calibration operations when damaged that can add cost, complexity and cycle time to each repair,” Mandell explains.

    While repairable EV claims increased, average claims severity for repairable BEVs fell by 2% in Canada (from $7,410 in 2024 to $7,253 in 2025) and 5% in the U.S., Mitchell reports. While remaining flat in both countries for PHEVs, claim costs for MHEVs increased by 4% to $5,054 in the U.S. and held steady in Canada at $6,267.

    The report also found total loss market values dropped across most powertrain types, with BEVs experiencing the largest decline of 13% in Canada and 6% in the U.S. “This was the result of accelerated depreciation, increased availability of lower-cost models and shifting consumer sentiment,” the release says.

    When it comes to the top North American BEV markets based on repairable claims frequency in 2025, British Columbia takes the lead at 8.48%, up 1% from 2024. Quebec comes in second at 8.21% (up 2.53% from 2024), followed by California at 6.58% (up 0.99% from 2024).

  • Do brokers trust their insurtech partners?

    Do brokers trust their insurtech partners?

    A mid-adult man and woman clad in hiking attire exchange a joyful high five on a mountaintop, silhouetted against a serene sunset sky.

    Trust is considered an essential part of any relationship in which one party is vulnerable to the actions of another. Insurtech is no different. And yet, the findings of the 2026 Benevolent Insurtech Trust Index reveal broker trust in insurtechs is not high.

    When two out of three brokers believe vendor claims of time-savings, efficiency and ROI are overstated, there is a trust gap.

    Consider the following findings in the inaugural Benevolent Insurtech Trust Index report, completed by 67 brokers from across Canada:

    • 67% of brokers surveyed say insurtech promises of time savings, efficiency and ROI (return on investment) are overstated;
    • 22% of respondents say vendors are honest about features, pricing and implementation during the sales process;
    • 23% of respondents say vendors can be counted on to do what is right;
    • 9% of respondents agree vendors have made sacrifices for them in the past.

    Before going further, a couple of observations about the survey. First, the sample size means results are indicative but not generalizable. Second, attitudes towards various categories of insurtech were used in the findings, including Broker Management Systems (BMS), Quoting/Rating, email marketing, Policy Admin Systems (PAS), and AI solutions.

    The survey suggests three areas in which trust is breaking down right now between brokers and insurtech vendors.

    To the moon and back: Promises of efficiency, ROI

    Perhaps the most challenging aspect of buying technology is validating vendors’ claims of ‘the better life’ post-purchase. Brokers, it appears, are less willing to accept claims of ROI at face value. They want proof.

    As one broker told the survey, “Show me real concrete examples of where our brokerage will see ROI and provide me with contacts that we could follow up with.”

    One possible cause of the ROI arms-race may be the fragmentation or specialization of insurtech. Every additional piece of tech puts pressure on vendors to help buyers justify the incremental spend. The problem isn’t necessarily the tech, but the expectation of outcomes.

    “Insurtech continues to offer strong potential, but the promises around automation, time savings, and efficiency don’t translate into meaningful improvements for frontline brokers,” one broker said.

    Also in the news: Do brokers or insurers own the customer relationship? A court may decide

    Other brokers expressed stronger feelings about the extent to which promises are embellished.

    “So yes, tech firms all overstate their ROI and what they can do for you because that’s how they get the sale,” said one broker in the survey.

    Which leads to the second trust gap that exists: honesty during the sales process.

    Why are your fingers crossed? Low trust in sales honesty

    Only 22% of survey respondents agree vendors are honest about features, pricing and implementation during the sales process. “Tech vendors in the insurance space suffer from the over-promise and under-deliver syndrome,” said one broker.

    Over-promise. Under-deliver. Overstated ROI to make a sale. Are these factually true?

    It doesn’t really matter. Even if it’s demonstrably untrue, what matters is the gap between trusting and not trusting the sales pitches. Right now, the gap is wide.

    How does this show up? Buying hesitancy. Longer sales cycles. And the ‘dark social’ conversations, those casual yet vulnerable chats between brokers at conventions and industry events where true feelings are shared.

    Me before thee: Vendors appear self-interested

    Consider these two findings: 23% of respondents say vendors can be counted on to do what is right. And only 9% of respondents agree vendors have made sacrifices for them in the past.

    Partnering relationships, those involving trust, are based on a willingness to be vulnerable to another party’s actions. Our survey findings suggest brokers are much less likely to allow themselves to be vulnerable to vendors. In fact, they expect vendors to act in a self-interested manner. This creates an environment in which information is less likely to be shared; and a vendor’s intent is less likely to be associated with the collaborative behaviour of a trustworthy partner. 

    How to increase trust

    Transparency in pricing. Realistic implementation timelines. Honest product roadmap discussions. These topped the list of ways brokers suggested vendors improve trust.

    “Trust grows with insurtech when [vendors] stop overselling roadmap features,” as one broker put it in the survey.

    Of course, being upfront about time savings, workflow efficiencies and ROI would also go a long way to strengthening feelings of trust. Those who are customer-facing have the biggest opportunity to break or build trust.

    The good thing is, trust can always be rebuilt, especially when goals are aligned. As one respondent stated: “Real insurtech success isn’t about disruption; it’s about reliability, partnership, and making brokers better at serving clients.”

  • Halifax council set to debate bylaw proposing increased oversight for Uber

    Halifax council set to debate bylaw proposing increased oversight for Uber

    The Uber app as seen on an iPhone

    HALIFAX – Halifax Regional Council is expected to begin debate today on a proposed bylaw requiring more oversight of drivers working for ride-hailing companies like Uber.

    If approved, rules for ride-hailing services would be brought in line with those covering taxi and limousine companies.

    The bylaw includes requirements for drivers to send the city results from training and background checks and an extra $135 in fees.

    Uber Canada is opposed to the proposed changes, saying Halifax already has the authority to request documents from Uber to determine whether drivers are complying with applicable laws.

    The company has said the proposed rules are redundant, full of red tape and would lead to higher fares.

    Debate was originally scheduled for two weeks ago, but the council pushed it to today.

    The proposed bylaw came out of a staff report presented to a council committee in December, and it’s unclear how many of the 16 councillors will vote for it. Mayor Andy Fillmore has come out against the proposed changes.

  • Are brokers getting squeezed by ChatGPT for the hearts and minds of consumers?

    Are brokers getting squeezed by ChatGPT for the hearts and minds of consumers?

    Digital warning icon with AI and data interface

    If your clients are relying on artificial intelligence (AI) searches to inform themselves about insurance, rather than professional advice from an insurance broker, that can be a risky proposition, warns Rates.ca.

    Anecdotally, the rate aggregator has seen an uptick in customers asking questions, making assumptions, or writing emails using AI or AI-generated information, says Rates.ca insurance expert and licensed insurance broker Daniel Ivans. And while AI can be a useful starting point for research, it’s important for the client to check their source of information.

    “AI often produces incorrect or out-of-context insurance information, or insurance information from different geographies,” Ivans says. “So, an Ontario driver may get advice for insurance in Nebraska. Or a homeowner in Alberta may get insurance advice that is applicable in Halifax.” 

    It’s important to note Canadian auto insurance policies vary widely according to geography, Ivans says. In addition to public versus private auto insurance models, “legalities, policy structures, endorsements and claims processes will differ according to each province. What applies to drivers in Ontario will not necessarily apply for drivers in Nova Scotia.”

    Similarly, home insurance advice is often based on local geographic and weather conditions. While home insurance policies don’t vary by province in the same way auto policies do, brokers’ advice is often based on the specifics of the home’s locality. For example, hail is a major risk in Alberta, but not as much in Ontario. Some localities also experience higher rates of break-ins or flooding, which can affect risk. 

    As well, brokers can offer tailored advice based on specific driver and home profiles. Auto and home insurance premiums and recommended endorsements can vary from person to person and from home to home. Brokers can provide detailed, tailored advice based on a variety of individual factors such as driving history, age, address, etc.

    Although AI can compile and summarize information quickly, it often does not discern between reputable sources of information and unconfirmed sources of information, Rates.ca warns. When using it to search for insurance information, clients should remember to check the source of the information provided and whether it’s reputable. Then check in with a broker before making any policy decisions.

    Brokers use of AI

    Meanwhile, Canadian P&C brokers are using AI as a core capability to place commercial and personal cyber and tech coverage, according to a new survey from BOXX Insurance.

    BOXX told CU the survey involved 92 brokers in Ontario. It conducted primary research among Ontario-based brokers offering tech and cyber insurance to understand AI usage and how brokers are engaging with AI tools in their day-to-day activities.

    The findings reveal AI is now embedded within everyday broker workflows, helping them simplify complex cyber risks, discover emerging threats, and deliver faster, clearer, and more tailored advice to clients. All cyber and tech brokers (100%) surveyed reported using AI in their day-to-day work at least multiple times per day.

    Among other key insights from the research:

    • 52% of brokers in the study say they use AI to explore trends, 47% to compare information, and 35% rely on AI tools or chatbots to stay informed on market shifts and emerging risks
    • Despite the surge in AI usage, Canadian brokers aren’t abandoning traditional wisdom. Brokers access sources like cybersecurity company reports (63%), industry publications and trade media (53%), cyber incident databases or data platforms (45%), social media and professional networks (43%), and professional associations or broker networks (41%) to understand industry trends, emerging risks and market developments.
    • ChatGPT is the dominant AI platform, used by 60% of brokers, far ahead of Google Gemini (12%) and Perplexity (7%)
    • Enhancing underwriting and risk assessment is seen as AI’s highest-value application (33%), followed by threat intelligence and real-time monitoring (15%), broker decision support (13%), market analysis (8%), and training and knowledge-sharing (7%)
    • Over half of brokers use AI to explore trends (52%), draft content (50%), compare information (47%) and simplify complex material through summaries or explanations (39%).

    Like Canadian brokers, the vast majority (94%) of Canadian CEOs are using AI to some extent, according to PwC Canada’s 29th Annual CEO Survey of 133 Canadian CEOs. But only 29% of Canadian CEOs report scaling AI across their business, compared to 43% globally, PwC Canada says in a release. “The winners will be those who act decisively now to embed AI enterprise-wide and unlock new revenue streams.”  

  • Will a polar vortex cause havoc for your commercial clients?

    Will a polar vortex cause havoc for your commercial clients?

    Frozen pipework at an industrial facility

    Canada’s cold right now, and businesses in this country are well-versed in property risks caused by freezing temperatures. Yet even a single freeze-related oversight can trigger multiple insured losses.

    For example, a failing chip in a policyholder’s heating, ventilation and air conditioning (HVAC) control panel can result in frigid air being blown into poorly insulated spaces. Water pipes in that space can freeze, then burst and flood offices – destroying servers, floors and furnishings.

    Likewise, frozen clumps of wood chips could jam an outdoor conveyor at a pulp and paper mill. Or a snowfall might melt and then re-freeze, crushing the roof of a factory while its staff is away on vacation.

    Increasing climate variability complicates freeze-risk management. Warm air in the Arctic makes North America colder by disrupting the polar vortex, sending sub-zero temperatures into hot places – as occurred in the deadly ‘Great Texas Freeze’ five years ago.

    Overcome complacency

    Freeze losses are surprisingly prevalent. A study our firm conducted on losses over a recent 20-year period finds our Canadian operation saw nearly 650 freeze-related losses totaling US$110.22 million.

    Industry is more effective at preventing freeze-losses than businesses with fewer workers on site. Roughly nine out of 10 claims we see are for facilities like offices, condominiums, hospitals, and schools.

    Even careful managers can become complacent about freeze risk, overlooking hazards such as:

    • old, untested heat-trace wire that no longer works at -40° C
    • wet-pipe sprinklers in areas subject to temperature change
    • underground water mains located less than the recommended burial depth and lacking heat tracing or pumps to keep water circulating

    Dry-pipe sprinkler systems can also lead to losses. For example, a riser room could lose heat, causing the dry-pipe valve to trip and fill with standing water. Also vulnerable are boilers, conveyors, and process equipment in warmer locations that can be exposed to the elements. Likewise, dryers for pressurized instrument air systems may be undersized, and even steam systems may freeze if condensate-return piping is not insulated.

    But if underwriters and brokers understand these hazards, they can help client businesses reduce loss.

    Oversight risk

    In places where frigid temperatures are normal, a typical freeze loss results from a change within a facility. This could be as simple as failing to replace insulation after a repair or leaving a door or window open. Sometimes, automated louvers (angled slats used for ventilation) fail to close, freezing nearby sprinklers and water coils in air-handling units.

    To remedy this, a continuously roving watch team of people should monitor building temperatures. This is particularly important for idle or vacant buildings, even if they’re only closed for a few days.

    In regions where freezing is rare, it’s important to have an alert weather watch and onsite portable heaters. Inadequate heat in stairwells and above suspended ceilings – along with inadequate insulation – can lead to freeze-ups. And frozen sprinkler systems can make the facility vulnerable to fire.

    In temperate regions, local infrastructure can be an issue. Even in Canada, some areas may not have enough snow removal and ice-treatment equipment to respond to a weather event spurred by a significant temperature shift. This can result in roads that are impassable for days.

    Utilities may also be reduced or lost, affecting production and heating. In these cases, it’s vital to identify equipment and piping that should be drained, and to train operators and maintenance staff how to prevent process upsets when draining.

    A winter checklist

    Here are few ways clients can get through challenging winters unscathed:

    1. Make a freeze-emergency plan that includes a reliable team to watch the weather. Before cold snaps arrive, revisit training so employees follow the plan.
    2. Install a backup fuel source if you have an interruptible contract that cannot be changed.
    3. Obtain portable heaters to use in cold weather emergencies.
    4. Keep operating during deep freezes but prepare for power outages and resulting loss of heat.
    5. Inspect facilities frequently, ensuring windows, doors, and other openings in the building envelope are closed. Make sure automatic louvers are operating and determine where the facility may be vulnerable.
    6. Establish preventative maintenance programs for the winter months.
    7. Test heating systems regularly, and secure ample fuel supplies.
    8. Inspect heat-tracing equipment in outdoor piping and equipment, and confirm insulation is in place.
    9. Verify dryers on instrument air systems can eliminate condensation during the coldest weather.
    10. Use freeze maps to spotlight areas where the 100-year return period daily minimum temperature is -6.7°C or colder.

    Even the smallest oversights can disrupt businesses that have weathered freezing temperatures for decades. Freeze losses are surprisingly common, and vigilance can keep your clients in business.

    Neil M. Chaudhuri is operations chief engineer for Canada at FM Global, a commercial property insurer.

  • What AI means for P&C talent development

    What AI means for P&C talent development

    AI robot recruiting employees

    As insurers move from artificial intelligence (AI) pilots to embedding AI into core underwriting and claims workflows, 2026 will be an important transition year for Canada’s property and casualty (P&C) insurance sector.

    AI is already delivering tangible benefits, including faster turnaround times in claims, underwriting and customer service, while improving efficiency and consistency across high-volume operations. 

    Less visible is how the shift will influence the way expertise is required to make critical judgments; and how that is developed inside insurance organizations.

    For decades, the P&C industry relied on a layered talent model. Entry-level employees performed routine and repetitive work, which over time exposed them to claims patterns, policy interpretation, and exceptions. This familiarity developed employees’ ability to make judgment calls. They acquired key skill sets through experience, repetition, and mentorship rather than through formal instruction.

    This structure supported continuity and helped maintain technical depth across generations of P&C industry professionals.

    As AI takes on more of this foundational work, senior expertise is not being displaced, but the pathway through which individuals develop into senior experts will narrow. By streamlining today’s workflows, insurers risk reducing the number of roles that historically supported the development of judgment.

    Regulation and efficiency

    Converging operational and regulatory changes are creating momentum behind this shift.

    New pay transparency requirements will soon mandate public disclosure of salary ranges. These measures are intended to promote fairness, consistency and accountability, in line with broader regulatory and societal expectations.

    Operationally, though, they may influence hiring behaviour.

    When compensation is transparent and hiring processes are more closely scrutinized, organizations often become more cautious when making recruitment decisions. A company’s flexibility can narrow. Developmental or longer-term hires may be more difficult to justify, particularly in operational environments where performance metrics emphasize speed, volume, and near-term outcomes.

    At the same time, AI is increasingly part of the hiring process; AI-enabled efficiency increasingly rewards immediate contribution. Together, these dynamics can shift hiring preferences toward candidates who arrive with fully formed experience rather than those expected to grow into a role.

    Over time, this may create a more polarized talent structure, with experienced professionals concentrated at senior levels, fewer entry points at early career stages and a thinner middle layer that has historically produced future technical leaders and managers.

    Time horizons

    Workforce risks rarely materialize quickly.

    The P&C insurance sector has long operated with an implicit time lag, as junior employees gain experience, mid-level professionals mature, and senior leaders eventually retire. Historically, this gradual progression has let organizations avoid significant disruptions.

    But the current environment is characterized by numerous pressures simultaneously affecting the talent pipeline. AI-driven efficiency is reducing early-career exposure to complex files. Regulatory expectations are shaping more conservative hiring practices. The remaining workforce continues to age. Simultaneously, fewer international students are entering Canada to supplement domestic talent pipelines in insurance-adjacent disciplines supporting underwriting, claims, analytics, actuarial, and operational roles.

    Although the implications of these trends may not manifest in the near term, they’re more likely to surface between 2030 and 2035. By then, the trend toward retirements will have accelerated and fewer experienced successors will be available to step into complex underwriting, claims and leadership roles.

    By the time gaps show up in underwriting outcomes, claims handling or leadership depth, remediation could be difficult. Rebuilding lost capabilities won’t happen quickly or easily. In this sense, the issue is less about short-term hiring and more about long-term structural resilience.

    Industry considerations

    The P&C industry has shown it’s capable of identifying and managing near-term operational risks. AI adoption is improving consistency and customer experience, while regulatory changes are reinforcing transparency and accountability around job descriptions and hiring practices.

    Key to moving forward will be the ability to match these advances with sufficient attention to how judgment and experience are cultivated over time.

    While efficiency gains are visible and measurable, changes in judgment capacity, if they happen, will likely be gradual and subtle, and only measurable after several years.

    Today’s gains support, not constrain, the industry’s future decision-making capacity to sustain performance over the coming decade.

    Jorrit Wit, is managing director and talent advisor at Fractional Recruiting.

  • How self-driving trucks could change Ontario’s insurance equation

    How self-driving trucks could change Ontario’s insurance equation

    Self-driving truck on a highway

    A major Canadian grocery chain is planning to put self-driving delivery trucks on the road in Toronto by year’s end, and an insurance expert tells Canadian Underwriter that may not be as much of a risk as people think.

    In a late-September press release, Loblaws says it signed a contract with Gatik, a U.S.-based startup. That deal would see 50 autonomous trucks delivering groceries and household goods to around 300 stores, according to news reports. 

    While many details on the insurance side will need to be sorted, there’s no need to panic, says Crystal Kaustinen, client executive for Commercial Insurance at Acera Insurance.

    She tells Canadian Underwriter that driverless vehicles operating on Ontario’s public roads must currently still have a driver in the cab as a safety measure. And Ontario has a close eye on the sensor systems in these trucks, she adds, noting they’ve demonstrated the ability to perform in several weather classifications, including ‘small snow,’ ‘small rain’ and ‘heavy rainstorms.’

    “They can’t do heavy snowfall. They can’t do quick stopping. It is going to be a problem, especially with that heavy load behind them,” Kaustinen says. “In Ontario, in the winter, our [highway] lines do deteriorate, and [in] construction zones there’s [new] lines everywhere.”

    Also in the news: How auto insurance will navigate a world with self-driving cars

    In terms of allowing driverless trucking, she says Ontario walks a middle line between U.S. states like Texas and California, which allow the vehicles, and European countries that apply far tighter restrictions. “It’s still pilot project. It is still very monitored by the government,” she tells CU.

    If these delivery vehicles are involved in accidents, she adds there will be inquiries. But they may not focus strictly on automotive insurance coverage.

    Instead, she says, they may look at issues like cyber security, the possibility of vehicles being taken over by outside actors, or an entire delivery fleet being taken down.

    In such cases, questions would examine “what has the fleet owner done to ensure the safety of these vehicles and of other people on the road,” Kaustinen tells CU. “[Or, for] the software vendor, did they make mistakes? Is there an error in [an] update?”

    Answers to such questions, she adds, will impact what requirements are set by governments via legislation to enact specific rules to ensure the safety of all road users.

    Coverage calculations

    For now, Kaustinen notes, auto liability is a shared responsibility and insurance coverages are based on historical data. But new autonomous driving technologies could alter the risk calculus due to the high value of sensors and radars on the trucks. “How much is the repair cost on these vehicles compared to traditional trucks?” she asks.

    Also, they could be affected by software bugs and problems stemming from large-scale shutdowns of communications networks.

    Those changes could mean fleet insurance for trucks might, at some point, become a “completely non-personal line [since there would be no] human being at the end of the end of food chain,” she says.

    In such cases, she says, new coverage models will have to respond to new categories of risks once the P&C industry has examined them.

    “Insurance already is not inexpensive. So, if you’re going to want to go this route, it’s going to cost more, especially [in the] beginning until we can get sorted,” she says. “We have to get…the data and know how it actually works.”

    Also in the news: Why the holidays spike truck accidents in prairie provinces

    And cyber is likely to be the biggest exposure.

    “Regulatory [and] insurance systems need to evolve alongside…the technology to ensure [people] stay safe,” she says.  

    “Failure of the sensors, [can happen if] it’s a weather-related, cyber-related, or data issue. But cyber products will [have to] cover those types of things. They’re not just for if somebody steals my data,” Kaustinen adds. “I haven’t seen a cyber product for trucking, manufacturing, or malfunctioning of a self-driving vehicle, specifically on commercial fleets. That [will] need to evolve before we can get launched.”

    She says the P&C industry has historically grown to meet new customer needs – and cyber coverage, which didn’t exist 15 years ago, is a prime example. “As this develops, the industry is going to have to grow and adapt to ensure that we’re able to cover these risks.”

  • Broker fined after tech startup advertises insurance before it was licensed to sell it

    Broker fined after tech startup advertises insurance before it was licensed to sell it

    Car insurance policy form on laptop screen. Insurance agent or salesman providing security document. People buying auto, leasing. Protection, warranty of vehicle from accident, damage or collision

    A B.C. broker who worked with a tech startup company as a side gig has been fined $2,000 after his start-up tech company went public with a website that advertised insurance services about a year before the company was licensed to sell insurance.

    Complicating matters for Barzin Assadi, the Insurance Council of B.C. found that he breached rules around a client’s privacy of personal information by emailing it to himself from the brokerage to his personal Gmail account.

    B.C.’s broker regulator noted Assadi had been disciplined for a privacy complaint in the past in B.C., which he did not disclose to the Manitoba regulator. When the Manitoba broker regulator disciplined him for not disclosing his disciplinary history in B.C., he did not subsequently tell the B.C. regulator about the Manitoba regulator’s decision.

    “Council considered mitigating and aggravating factors,” the Insurance Council of B.C. wrote in a decision released in October 2025. “Mitigating factors included that the [broker] had co-operated with council’s investigation, that clients were not harmed by [making the website public prematurely], and – most significantly – that the [start-up’s] website had been shut down very quickly when the [broker] had been made aware that it was publicly accessible.”

    However, council was concerned about the broker’s handling clients’ personal information in the past and not disclosing his disciplinary history to other provincial regulators.

    “[T]here is, in council’s opinion, some likelihood that the [broker] may repeat the misconduct relating to client confidentiality and notification of regulators if disciplinary action is not taken,” the ruling states.

    Premature website

    Council found Assadi, while still working for his former brokerage, became involved with a startup technology company that was incorporated in 2020. At that time, Assadi told council, it was never intended the technology company would sell insurance or other products.

    But when Assadi and his colleagues at the tech start-up eventually decided to sell insurance, they created a brokerage, which would become licensed to sell insurance in April 2023.

    Before then, in June 2022, and again in November 2022, council noted the website was accessible by the public. In June 2022, council reviewed the brokerage’s website, making the following observations, as cited in the regulator’s decision:

    • The website contained statements such as, “Finally, a fast and easy way to get the best home insurance coverage in BC. 90 seconds is all it takes.”
    • The website described how AI could help provide simple and affordable home insurance coverage, and that “Kylie,” an AI agent, could show clients the best insurance rate options.
    • The website described the agency as an “online insurance brokerage” and noted that its AI could provide quotes, customize policies, and activate coverage.
    • The website stated: “Using powerful Artificial Intelligence (AI) technology, [the Agency] revolutionizes the home insurance buying process.”
    • The website contained a testimonial from a purported customer that stated: “Wow, someone finally figured out how to make this process not suck.”
    • The website allowed for completion of an application, up until the point of being asked to provide payment details.
    • The website featured headshots and names of the agency’s leadership, including Assadi as the president.
    • The “Terms and Conditions” portion of the website appeared to have been copied and pasted from the website of an unrelated insurance agency.

    Assadi told council the website had not been “operational” during the time it was made public before the brokerage was licensed to sell insurance.

    “’Not operational,’ in the sense intended by [Assadi], meant that the website was under construction and that [insurance] applications could not be completed or finalized,” as the council decision reads, adding the broker wasn’t denying it had been made public.

    “It was submitted that there was no payment vendor set up for the agency website at that point, and that it would have been impossible for an application to be completed and paid for. The first full transaction completed on the website occurred in November 2023, after the agency had become licensed with council.”

    Council’s investigation corroborated the story, finding the website traffic was low during the times it was public. Plus, the geographic locations of the website hits corresponded to those of the staff members designing the site.

    As for the testimonials and terms and conditions, those were mere “placeholders” until the time the website was ready to be released.

    Council granted members of the public were not harmed by the premature public presentation of the website, but it nevertheless held the public release of the website did breach the rules.

    “Although it was in error that the agency website was made accessible to the public, council’s opinion is that [Assadi] should have taken greater care to ensure that the unlicensed company he had a prominent leadership role in was not improperly advertising itself as a provider of insurance services, or otherwise publishing misleading marketing information,” council ruled.

    Client confidentiality

    One aggravating factor in its sentencing, council said, is that Assadi had emailed client information to himself at his own personal Gmail account.

    Assadi argued the information wasn’t particularly sensitive. He acknowledged two incidents in which he sent client information from his former brokerage to himself.

    “When asked by the committee as to how he had inadvertently sent the same client’s information to his Gmail account on two separate occasions, [Assadi] stated that he may have been attempting to send them to another representative of the former employer but that the auto-populate feature may have entered his own Gmail address rather than the intended address,” council’s decision states.

    B.C.’s council had investigated Assadi previously for how he handled client information. It issued a disciplinary order in September 2021 “for client confidentiality issues similar to those investigated in the present case,” B.C. council noted.

    Assadi was subsequently disciplined in June 2023 by Manitoba’s broker regulator for failing to disclose the B.C. investigation.

    B.C.’s council subsequently found he did not disclose Manitoba’s ruling to them.

    Assadi’s lawyer said his client had intended to provide the B.C. council with notice of his Manitoban discipline “after receiving a payment acknowledgment letter from the [Manitoba broker regulator],”

    But B.C.’s council found the five-day rule for notification had clearly been breached.

  • Using AI to assist rather than replace your workforce

    Using AI to assist rather than replace your workforce

    Woman and robot working together

    Fears of artificial intelligence (AI) replacing human workers is a common sentiment in the insurance industry and other sectors, but one firm is focusing on how AI can augment its processes rather than take them away.

    “We ask how AI can play a role in any project that we launched, whether it’s a large-scale project or tiny project,” says Sébastien Mc Mahon, chief strategist, senior economist, portfolio manager and vice president of asset allocation at iA Global Asset Management (iAGAM). “We need to ask ourselves the question, ‘What can I do with AI here?’”

    iA has teams drumming up ideas to integrate AI into “any kind of process,” Mc Mahon says. For example, “help for call centres — having an AI listen in to a conversation and being your wingman rather than completely replacing the person on the call.”

    AI can also be used to listen in to investment conversations and opine on any blind spots due to groupthink, Mc Mahon says during KPMG’s 2025 Insurance Conference last week.

    “We like to call it augmented intelligence rather than artificial intelligence for now,” he says. “At some point, likely the artificial part will kick in, but now it’s more augmented.”

    iA is also using language models to create custom indicators about how sell-side research (from big banks and big research firms, for example) evolves over time. “Who should we pay more attention to, in which kind of environment?” Mc Mahon asks. “These are language models, so they’re good with inferring from language. They’re pretty good [at] dealing with this data, the algorithms that we have now.”

    For Co-operators, the insurer has put its data in the cloud and created digital solutions for its clients, particularly on the property and casualty side of the business, reports Mary-Jane Jackson, the insurer’s vice president of enterprise transformation. “A lot of our business historically was through our advisor offices, and so we needed some digital solutions to be able to ensure we’re meeting clients where they need to be met.”

    Co-operators has a ‘steering committee’ of sorts to look at and approve any AI applications, Jackson reports. It’s currently looking at opportunities to take written data and translate it into health forms and documentation, among other uses.

    Regulatory impacts

    The impacts of AI continue to evolve, and regulators have been trying to address emerging technologies for years. For example, the Financial Services Regulatory Authority of Ontario offers a regulatory ‘sandbox’ that allows the industry to test out new ideas, concepts and innovations.

    Mc Mahon suggests the industry try to get ahead of AI by working with regulators proactively.

    “One bold move could be to make sure that there’s an industry front; just move to the forefront of the AI adoption by working right away with the regulator,” he says. This would accelerate innovation, adoption of the technology, while maintaining governance and consumer protection.

    “You can build a sandbox,” he says. “The industry works with your regulator [to] try to tackle everything upfront by including all of those aspects, rather than having things that are decentralized and then having the regulator having to juggle all of these balls.

    “That would be a smart thing to do as an industry; to be at the forefront.”

  • One industry pro’s trying experience starting a claim

    One industry pro’s trying experience starting a claim

    Man talking on the phone at home

    Canada’s insurance industry could use emerging technologies more effectively to give customers a ‘first notice of loss’ (FNOL) experience they’ve come to expect in other industries, attendees at KPMG’s 2025 Insurance Conference heard Tuesday.

    Tom Reid, digital experience lead with the Insurance Brokers Association of Canada, shared his experience using technology to successfully handle a plumbing situation compared to waiting hours on the phone trying to start a hail claim.

    Reid says he realizes companies cannot be staffed 24/7, but his FNOL experience with his insurer left much to be desired.

    “Both companies…[are] super well aware that when a customer is calling them, they have a problem,” Reid says during a panel discussion on innovation and the adoption of emerging technologies. “They need that problem resolved as soon as humanly possible.

    “And what’s different with these two stories is that that plumbing company made an investment in technology, updated the processes, so it makes me really stand out from the crowd…” Reid says. “On the other hand, my insurer, I was not quite as pleased with, and I instructed my broker to shop that policy.”

    Reid’s insurance experience occurred during a massive hailstorm in Calgary last August. After the storm, at about 8:15 p.m. one night, he went outside to discover damage to his house and car. So, he called his insurer’s 1-800 line and got their automated phone system.

    “So, it’s ‘press one for a car claim, press two for a property claim,’” Reid says. “Well, where’s three, both?”

    Reid ended up using his own phone for the auto claim and his wife’s phone for the property claim. “And waited and waited…I heard their privacy policy…I could actually verbatim repeat the privacy policy by the time I was completely annoyed here.”

    At about midnight, Reid hung up. He remembered his insurer had a FNOL form on their website — “essentially, it’s a ‘contact us’ form,” he said — so he entered his contact information. “Good news, the next day, I did get a call from the auto adjuster. About three days later…the property adjuster.”

    Different experience

    A couple of months later, Reid’s daughter-in-law discovered water under ceramic tiles in an ensuite bathroom. Reid used his insurance background to help determine the water was coming from the toilet, which was turned off.

    At about 8:30 or 9 p.m. one night, he called the plumber, as his extended family is under his service contract.

    “Much to my surprise, the phone was answered by an AI receptionist, which identified itself as AI,” Reid says. “It knew who I was already, because I had my phone number in the system, and it says, ‘Are you calling about this house or that house?’

    “We actually talked back and forth and…[it] ascertained this wasn’t an emergency,” which would have been expensive, Reid says. “The AI said, ‘Hey, I’m going to take notes of this and leave a note for my human counterparts, and somebody will get back to you.’”

    The plumbing shop opens at 8 a.m., and three minutes after opening, Reid got a phone call from somebody at the shop saying they received all the notes from the AI receptionist.

    “‘So, we know exactly what’s going on. We have a plumber en route, and he’ll be there in 30 to 45 minutes,’” Reid says. “I was pretty blown away.”

    While the insurance industry has come a long way, consumers still want automation and real-time service, Reid says. For example, utility services provide power outage maps and wealth management platforms provide 24/7 access and retirement calculators. “If a little 50-person plumbing company in Calgary can do it, so can we,” Reid says of Canada’s P&C industry.

    Comparing buying decisions

    “Consumers make hundreds — maybe thousands in some cases — of buying decisions every year,” he says. “One or two of those is an insurance transaction. So, their comparison is not insurance company A to insurance company B; it’s my insurance company or my potential insurance company, versus all the other decisions that I’m making today…”

    Insurance and other highly regulated industries can have challenges innovating within their existing environments, acknowledges Karthik Ramakrishnan, CEO of Armilla AI. But as a startup, Armilla AI doesn’t have encumbrances of a large company, such as legacy systems.

    “Large companies are very embarrassed to fail,” Ramakrishnan says. “They don’t want to put out a product that fails at the first factor.

    “Do you know what the motto of a startup is? ‘If you are not embarrassed in the first version of a product, you did something wrong.’ That is how we approach it.”