Beyond Demographics: The Benefits of Data-Driven Segmentation in Insurance
Tech-savvy insurance companies are already using advanced segmentation techniques to tailor customer interactions.
Building-Based Geocoding Helps P&C Insurers Assess Climate Change Risk
As the frequency and severity of natural disasters rises due to climate change, property and casualty insurers are facing immense risk when it comes to insuring property in vulnerable areas. To better assess and calculate climate risk in a changing world, some insurers are turning to a geospatial information system (GIS) functionality called geocoding.
The Role of Driving Simulators in Risk Calculations in Auto Insurance
Driving simulators aren’t new to the auto industry. From testing new vehicles’ performance and design to evaluating automated driving technologies and driver assistance systems, vehicle manufacturers have long leveraged simulators to test drive vehicles without the need to construct physical prototypes, saving time and money.
The 2024 Starting Line is Fast Approaching for Insurance Companies
As 2024 approaches, aligning resources with technical and operational needs is critical for insurance companies.
The Promise (and Reality) of Blockchain Adoption in Reinsurance
Today, the reinsurance process is rife with tedious manual work, one-off-contracts, and months of back-and-forth communication before contract signings are complete. Insurers use multiple reinsurers, and data exchange and information sharing can be inconsistent depending on the organization. According to CB Insights, blockchain technology can save reinsurers up to $10 billion by cutting costs through automation and reducing risk by facilitating information sharing. A PwC report found that reinsurers are building some of the largest blockchain applications outside of payments, opening a $5-10 billion cost savings opportunity due to “faster, more efficient, and more accurate placement, claims settlement, and compliance checks such as sanctions.” The promise of blockchain technology has already begun making waves in reinsurance, sparking technological developments that aim to automate parts of the reinsurance process. More specifically, smart contracts have emerged as one way to help reinsurers automate and streamline many pieces along a claim’s journey. Automating Reinsurance with Smart Contracts Securing reinsurance contracts through smart contracts is one of the ways that reinsurers are beginning to leverage blockchain technology to simplify claims processes and information exchange. Smart contracts, or programs stored on a blockchain that automatically run when pre-determined conditions are met, can be used to automate claims processes based on triggers. When reinsurance contracts are secured on the blockchain through smart contracts, the exchange of information between insurers and reinsurers can be simplified. Transactions pertaining to premiums and losses can be updated on an insurer’s and reinsurer’s systems simultaneously, removing the need to reconcile books between institutions for every claim. Duplicate manual work can be eliminated. Smart contracts are particularly helpful when it comes to parametric insurance contracts, or insurance contracts that automatically pay out a pre-agreed amount when specific events occur. For example, an agreement can be encoded in a smart contract to pay out $10 million if a category 4 or 5 hurricane occurs in a defined area. A good example that highlights both the promise and the reality of blockchain implementation in reinsurance is the story of B3i, an insurance industry consortium founded in 2016 by some of the world’s biggest insurers and reinsurers to explore the use of blockchain technology in the industry. The consortium began with insurers Aegon, Allianz, Munich Re, Swiss RE, and Zurich, who were joined by 10 other insurers in 2017. Among B3i’s endeavors included the launching of a prototype of a smart contract management system for Property Cat XOL contracts, a type of reinsurance for catastrophe insurance. B3i’s smart contract automatically evaluated data sources and calculated payouts after events such as hurricanes or earthquakes occurred. While B3i successfully raised tens of millions in funding in its early years, it was forced to file for insolvency due to lack of funding in 2022, despite backing from some of the insurance industry’s biggest names. B3i’s bankruptcy brought with it questions about the viability of blockchain as a platform in the reinsurance industry. But a few other market players have already emerged in its place. The first blockchain-powered reinsurer, Re, launched in late 2022. Alternative capital providers back baskets of insurance policies to earn premiums and yields through Re, whose smart contract protocol is built upon the Avalanche blockchain. In June 2023, the global climate risk coverage platform Arbol partnered with The Institutes RiskStream Collaborative to launch a blockchain-powered parametric reinsurance platform called dRe. dRe is a smart contract-based system that leverages validated weather data from decentralized networks to trigger smart contracts based on wind speed and location for peril events. Claim initiation, notifications, and loss calculations are all automated in the system, speeding up payouts. While the potential for smart contracts in the reinsurance industry is clear, their full potential in the space has yet to be reached. There’s no doubt that the future will bring more exciting developments and additional market players eager to explore the benefits of smart contracts in reinsurance.
How Life Insurance Companies Can Use Tech to Reduce Fraud
How life insurance companies can implement technology to combat fraud in life insurance applications and claims, and explain how fraud prevention can improve efficiency and the bottom line.
P&C Insurers with Aerial Imagery Technology
In July 2023, a California homeowner made headlines after his longtime home insurance carrier terminated his policy. He’s not the only one—California is facing an increase of policy non-renewals as insurers leave the state due to wildfire risk. However, disaster risk may not have been the reason this particular homeowner lost his policy. What was headline-making about this case was that the homeowner was told by the insurance company, the California State Automobile Association (CSAA) Insurance Group, that the decision was made due to aerial photos taken of his property. Those photos showed hazards in his yard, so he lost his coverage. “Apparently [the insurance company has] some [aerial] pictures, and they noticed clutter,” Oakland, California resident CJ Sveen told a local ABC news station. His story is just one example that highlights how aerial imagery—any image taken from an airborne craft like a drone, balloon, or airplane—is becoming more prevalent in P&C insurance. More broadly, the global market for aerial imaging is anticipated to increase to $8.1 billion by 2030, according to the analytics firm Research and Markets. As market growth continues, P&C insurers are increasingly using aerial imaging technology to determine both the legitimacy of insurance claims and the extent that properties are damaged. In doing so, they’re better able to flag fraudulent catastrophe (CAT) and not-CAT claims and more effectively move legitimate ones through the claims process. Aerial Imagery in P&C Insurance Aerial imagery has been used in insurance for years, with Deloitte noting that some leading insurance companies secured Federal Aviation Administration (FAA) permission to use drones for aerial data collection as early as 2015. Insurtechs have also emerged in the space to offer high-resolution aerial imagery to insurers. The aerial technology company Nearmap captures high-resolution aerial imagery of 100 million U.S. parcels three times a year, enabling insurers to compare historical data. Verisk captures high-resolution aerial imagery and augments it with analytics to reveal property risk. EagleView captures before and after aerial imagery of properties impacted by storms to help insurers expedite CAT claims. Those are just a few examples. Gaining a Clear Picture During Claims Processing With growing threats such as climate change, P&C insurers need to stay up-to-date with the most effective ways to assess risk. Gaining a bird’s-eye view into properties with aerial imagery can be one piece of the puzzle when it comes to better managing these threats. When disaster strikes, timing is critical. At the same time, it can be difficult—sometimes impossible—to physically enter and inspect an area that was recently ravaged by a natural disaster. With post-disaster aerial images, insurers can more rapidly assess damage and quickly process an influx of claims. More specifically, claims adjusters can use precise aerial imagery as part of their damage assessment. Was a home destroyed by a hurricane? Or is it just missing a few shingles off the roof? Aerial imagery can help answer those questions. Enabling Better Fraud Detection Another key use case of aerial imagery in P&C insurance is fraudulent claim detection. This is critical, as a study from the Coalition Against Insurance Fraud recently determined that total losses due to insurance fraud in the U.S. is at least $308.6 billion every year. If an insurer has access to historical aerial imagery, they can determine whether or not a change occurred to a property because of a natural disaster. They can promptly assess the imagery, gaining a clearer picture of the damage at an exact moment in time. This makes it possible for insurers to help prevent fraudulent claims for damages that actually occurred before or after the natural disaster. It’s not just effective in cases of natural disasters. One insurer partnered with Nearmap to pull high-resolution imagery of nearly one million properties in California to identify 80,000 pools that had been built between 2019 and 2021. Most of the property owners building these pools did not disclose those upgrades to their insurers, meaning they weren’t factored into premium rates. The potential for aerial imagery to help insurers expedite claims, gain better risk profiles, and mitigate fraud is clear, especially when combined with other forms of GIS-powered location intelligence. P&C insurers embracing this cutting-edge technology will likely see benefits to operational efficiency, underwriting and fraud detection.
Why Colorado’s Life Insurance Regulation is an AI Governance Game-Changer
Colorado continues to make waves as it moves forward with a first-of-its-kind U.S. law regulating how life insurers operating in the state (including many major national insurers) are allowed to use AI and big data. More specifically, the proposed regulation is designed to ensure that life insurers using external consumer data and information sources (ECDIS), algorithms, and predictive models don’t do so in a way that results in unfair racial discrimination. AI and predictive analytics are already proving their worth in the insurance industry, with more industry adoption expected. One Willis Towers Watson study found that 60 percent of life insurers say insights from predictive analytics have increased sales and profitability. Meanwhile, a report by McKinsey & Company has found that 10 to 55 percent of major insurance functions could be automated by 2030, including underwriting, actuarial, claims, finance, and operations functions. But AI bias has made its mark, too. There have already been instances of unconscious biases in algorithms going too far. For example, A.I based mortgage lenders are significantly more likely to reject Black, Latino, and Indigenous mortgage applicants than white applicants with similar financial characteristics. While the law hasn’t been finalized, it’s not too early for insurers to start preparing for when it does go into effect. This regulation is coming, and it has game-changing implications for AI governance in the insurance industry. A Closer Look at the Proposed Regulation In 2021, the Colorado Division of Insurance (DOI) passed SB21-169 – Protecting Consumers from Unfair Discrimination in Insurance Practices. According to the DOI, the “legislation holds insurers accountable for testing their big data systems – including external consumer data and information sources, algorithms, and predictive models, to ensure they are not unfairly discriminating against consumers on the basis of a protected class.” Two years later, the DOI followed up with more specific guidance to enforce the law, issuing the first draft of its proposed Algorithm and Predictive Model Governance Regulation in February 2023. After industry and stakeholder feedback, it released a scaled back version of its proposed new guidance in May. Even scaled back, the guidance imposes substantial requirements on life insurers. Guiding principles, board oversight, senior management accountability, a cross-functional governance group, written policies and processes, processes for addressing consumer complaints and inquiries, risk assessments and prioritization, and vendor risk management are all covered in detail in the revised draft as requirements. There are also significant documentation requirements, including a report summarizing compliance within one year of the revised regulation going into effect. The entire revised draft can be read here. Significantly, the law also places the responsibility of compliance on insurers in cases where they’re using third-party vendors. The revised document notes that if “an insurer uses third-party vendors and other external resources with respect to ECDIS as well as algorithms and predictive models that use ECDIS, the insurer remains responsible for ensuring all regulatory requirements are met.” Insurers also have to establish a process to select and oversee external resources and third-party vendors. Deloitte recommends that life insurers waiting for the final regulation to be issued can start by conducting a diagnostic assessment to identify and inventory AI use cases, develop an AI governance strategy, and train employees who work with ECDIS and AI models to understand new regulatory requirements. The Bigger Picture While the regulation in Colorado is the first-of-its-kind, it’s unlikely to be the last. The AI regulatory landscape is evolving in multiple states, making it probable that this regulation will be one among many that attempts to provide clarity into how insurers are allowed to use predictive models and AI. Additionally, laws like these will not stay confined to the life insurance industry. The DOI in Colorado has already stated in a stakeholder meeting that the scope of the regulation, or a similar regulation, is likely to widen to other insurance areas like auto, property, and casualty. When it comes to working with third-party insurtechs that offer AI or predictive modeling capabilities, insurers will need to tread carefully to ensure compliance. This could have implications for insurtechs as well, as insurers will come to them with higher expectations.
Shift Technology – Improving the Combined Ratio with AI
To help us understand how insurers are applying AI to improved the combined ration, we invited Jeremy Jawish, CEO of Shift Technology to join the FintechWire Innovators Podcast. Also joining us is insurance conlsultant Jeff Goldberg.
Generative AI Will Be Transformative for the Insurance Industry
Generative AI is in the spotlight, in large part due to OpenAI’s ChatGPT becoming a nearly instant phenomenon after its November 2022 launch. Industries ranging from banking and securities to education and real estate are exploring how generative AI solutions — artificial intelligence systems capable of generating original content — can streamline processes and offer additional value. The insurance industry is no exception. While the explosion of public interest in generative AI came on quickly, it’s no fad. According to a Bloomberg Intelligence report, the generative AI industry is set to become a $1.3 trillion market in revenue by 2032, up from $40 billion in 2022. Meanwhile, McKinsey estimates that generative AI could contribute roughly $50-70 billion in additional value for the insurance industry by boosting performance in functions such as marketing and sales, customer operations, and supply chain and operations. So, how can insurance companies leverage generative AI to achieve this additional value? Insurance companies, insurtechs, and software companies have already started to answer this question by exploring what’s possible with the new technology. Opportunities and Challenges Generative AI tools aren’t ready to fully replace the human touch of insurance claim handlers, underwriters, and customer service representatives. Even so, they do have the power to drive digital transformation in the insurance industry. With ChatGPT’s ability to generate natural-sounding content in seconds thanks to its large language model (LLM), there’s a lot of talk about its potential to streamline customer service, underwriting, claims process, fraud detection, and even risk management in the insurance industry. For example, ChatGPT could be used to automate the extraction and categorization of information from claims forms, which is typically done manually. “If we drill down to property and casualty specifically, claims have really bubbled up as an area where our clients are seeing a lot of potential opportunity. In the short term, there’s a significant amount of efficiency that can be gained through summarizing and synthesizing the significant amount of content that comes through a typical claims cycle,” explained Chris Raimondo, Americas Insurance Technology Leader at insurance consultant EY, in a recent webcast that explored the future of insurance powered by generative AI. Some insurance companies jumped on the opportunity to experiment with ChatGPT shortly after its release. The Swiss insurance company Zurich made headlines in March for “experimenting with ChatGPT to find out how AI can help with tasks including modeling, claims, and data mining.” Its experiment included inputting claims data from the past six years to determine the cause of losses across a large number of claims. At the same time, using ChatGPT raises security concerns around violating data protection requirements. These concerns come amid OpenAI’s regulatory scrutiny, including scrutiny from EU countries under GDPR and a FTC investigation into OpenAI’s data security practices stemming in part from a payment-related information leak. Another barrier preventing insurers from adopting generative AI tools like ChatGPT is their lack of AI maturity. A recent Accenture report that graded the world’s largest companies on AI maturity found that among 54 insurance companies, none had achieved AI maturity, which requires mastering key capabilities in both data and AI and organizational strategy, talent, and culture. Consumers themselves may have mixed opinions about using generative AI along their insurance journeys. A survey by the software company InRule Technology found that 70 percent of consumers still prefer to interact with a human over generative AI tools. These preferences differ drastically by age, with 71 percent of Boomers being uninterested in chatbots like ChatGPT versus just 25 percent of Gen Zers. Insurers considering generative AI for improved customer service will need to keep their range of consumers’ preferences in mind before replacing key human interactions. The Path Forward Despite these challenges, generative AI is making waves in insurance due to its breadth of potential use cases. A Sprout.ai survey of over 100 US and UK insurers found that 35 percent of insurance companies are already using generative AI in their processes. The AI insurance company Lemonade released a company shareholder letter in May mentioning plans to use generative AI for up to 100 of its business processes. Additionally, insurtechs and tech companies have already jumped at the opportunity to bring generative AI solutions to the market that target the insurance industry: As more insurers explore how they can leverage generative AI and more vendors come to market, its true impact on the insurance industry will become clear.