McKinsey predicts that underwriting, as we know it today, will cease to exist by 2030 – at least, for most personal and small-business products in life, property and casualty insurance. The processes around risk assessment and policy writing are largely still manual, tedious, and prone to human error, but this is set to change as insurers ramp up their investments in AI. In this article, you will find out how the acceleration of AI technology is transforming the underwriting industry and redesigning the role of human underwriters.
Insurance Is Experiencing a Digital Revolution
AI-driven digital transformation has had a profound impact on almost all industries, including banking, healthcare and manufacturing. While the insurance sector wasn’t the fastest to embrace AI technology, COVID and the emergence of agile, data-driven Insurtech startups have accelerated the adoption of AI. According to Gartner, over a third of insurers have deployed AI, and a further 16% are planning to invest in the next year. While there are many areas where insurers can apply AI to improve business results, underwriting is in the spotlight, as it is a core business function with a direct impact on customer experience, profitability and competitiveness.
Underwriting and the Urgent Need To Modernize
Despite some advances in automation in recent years, insurance underwriting is still a heavily manual process that is time-consuming and prone to human error. The lack of efficiency on the back end is a problem because customers today are used to ordering a product online in two clicks. Consumers across all demographics, from boomers to Gen Z, expect a fast and simple experience when they take out a life insurance policy or insure their new home. While a cool website and social media presence are great for engaging customers, they quickly lose their appeal if the underwriting process leaves the customer waiting for weeks to find out if they can be insured and under which terms. As business models evolve, competition increases and data sources multiply, it is clear that insurers need to optimize underwriting processes to reduce costs and increase efficiency. Gartner found out that almost one-third of life and pension underwriting has been at least partially automated, primarily to reduce policy issue time and cost.
AI vs. Traditional Underwriting
How exactly does AI help automate and optimize underwriting? Here are a few of the key ways that insurers are using AI technology to streamline their underwriting processes.
Improved Data Processing
Insurance companies have always been highly data-driven, as access to key information plays an essential role in accurate risk assessment. With the abundance of data available in today’s cloud-first and IoT-powered world, insurers can tap into a variety of data sources to build a more complete picture of the risks. For instance, an insurer could look at telematic data for car insurance, satellite imagery and weather data for home insurance, or fitness trackers for health insurance. Combined with more traditional sources of information, the data gives better insight into the applicant’s risk profile.
Making sense of this mound of structured, semi-structured and unstructured data is quite a task. Even classic sources of data, like contracts and other documents, can be time-consuming to process due to their unstructured nature. Insurers can leverage technologies based on machine learning (ML) and natural language processing (NLP) to parse the data faster and more accurately and expedite the underwriting process. Cortical.io SemanticPro uses Natural Language Understanding (NLU) to process complex documents more efficiently by automatically identifying key terms and conditions, even when the language is ambiguous, and the terms aren’t standardized, thus streamlining the underwriting process for insurers.
Incorporating AI into the underwriting process does not have to be a massive overhaul of infrastructure that takes years to roll out. With Cortical.io SemanticPro, insurers can quickly implement a custom AI solution and start generating actionable insights within weeks, limiting their risk and investment. This approach enables insurers to better understand their data and exponentially improve their underwriting workflows.
Better Decision Making
Human bias and inconsistency are more severe than a lot of insurance companies realize. In one study, researchers compared how different underwriters assess risk and price policies. The executives at one insurance company expected the difference between two underwriters to be around 10%, which they deemed acceptable. In reality, when the researchers submitted the same cases to multiple underwriters, there was a 55% difference in opinion. Human underwriters can use AI tools to support and verify their conclusions, which helps reduce bias and drive more consistent, fairer risk assessments and policies.
Reducing Human Error
Human error during manual underwriting can lead to financial consequences, missed opportunities, and fraudulent claims later down the line. Oftentimes, these errors stem from bias, insufficient or inaccurate information, or simply fatigue. We all make mistakes, but errors are costly for insurance companies. Many elements of underwriting require manual data entry and require high levels of concentration – like analyzing loss-run reports. This leads to fatigue and frustration, resulting in costly mistakes and sometimes even burnout. According to an SMA report, underwriters are frustrated by tedious, time-consuming tasks that can be performed by entry-level workers. Instead, they would prefer to focus on complex cases and high-level decision-making. One US-based insurer found that 70% of its policies still contained errors even after a manual review. Learn how they used AI to automate and standardize their global policies, improving accuracy and reducing manual effort.
Advanced Risk Analysis
Underwriters today are facing increasingly complex risk assessments, onset by the pandemic, war, and the climate crisis. As the complexity of risks increases, AI will play an essential role in helping underwriters to interpret relevant information buried in long documents or produced by connected devices. AI systems rely on historical data, so they aren’t a crystal ball. Predicting so-called black swan events – unpredictable and rare occurrences, like the start of World War I, or the outbreak of Covid 19 – is near impossible for anyone, human or machine. Nonetheless, AI technologies do excel at spotting patterns and predicting future trends. Global warming, for instance, is producing a whole new set of risks, and many of its effects should be predictable for advanced AI systems. An AI can analyze location and weather data within seconds, generating insights and future predictions in an easy-to-understand format for an underwriter. This is a much more likely scenario than expecting an underwriter to conduct a manual analysis, which could take months.
So, Will AI Fully Automate Underwriting by 2030?
Using AI, underwriters can make better decisions, be more efficient, and make fewer errors. Could AI automate underwriting entirely, cutting out the need for human underwriters? In a word, no. However, underwriters who use AI will outperform those who don’t. As human and AI collaboration becomes more commonplace, insurers will be able to improve operational efficiency while driving customer experience and improving employee satisfaction. Insurers that are taking steps to integrate AI into their processes, like underwriting, policy reviews and quotes, will be able to achieve a real competitive advantage as the number of connected devices increases and the complexity of new risks, like climate change, evolves.
Interested in learning more about how your company can use AI for underwriting and other insurance workflows? Book a free consultation with one of our specialist advisors or read our free whitepaper about intelligent document processing.
