Overview
From utilizing weather tools that aid in fortifying homes to employing an ecosystem of virtual agents, insurers have put aside their initial reluctance and welcomed artificial intelligence to help customers and better understand risk, according to Best’s Review. Beinsure analyzed the report and highlighted the key points.
Key Highlights on AI Adoption in Insurance
- Insurers use AI to automate underwriting, claims processing, and document comparison, allowing staff to focus on decision-making and risk evaluation rather than manual tasks.
- Companies like Coalition apply Generative AI to translate complex cybersecurity and coverage information into clear language, helping clients better understand risks and insurance terms.
- Flow Specialty uses agentic AI trained on insurance documents and broker interactions to handle quoting, risk matching, and document review, improving support for underserved markets.
- Liberty Mutual and Coalition deploy AI copilots to process broker submissions and assist employees with data extraction and formatting, improving speed and consistency across workflows.
- ACORD’s AI-powered tool processes over 4,000 insurance document types, extracting structured and unstructured data with high accuracy, supporting risk assessment and quote generation.
Insurers Accelerate AI Adoption to Improve Efficiency
Insurers are expanding the use of artificial intelligence across operations as confidence in its efficiency and accuracy grows. AI is increasingly viewed as a practical tool for reducing workloads and supporting staff functions, especially as advancements accelerate.
Edin Imsirovic, director at AM Best, noted during the firm’s Europe Insurance Market & Methodology Briefings in London that AI is becoming more advanced and accessible, with development moving at an exponential pace.
Despite regulatory concerns and uncertainty around implementation, several insurers and brokers are already deploying AI in production environments.
AI’s getting much better, much more useful and much more sophisticated. At the same time, the development of AI is accelerating exponentially.
Edin Imsirovic, director, AM Best
Commercial brokerage Flow Specialty has introduced a network of AI agents that passed industry certification exams and act as high-functioning support tools. Liberty Mutual developed an AI platform that offers weather preparedness advice to homeowners, factoring in property-specific risks and geographic data.
Zurich North America recently appointed a chief data officer to lead the integration of AI across internal systems, demonstrating a strategic commitment to AI-driven transformation.
ACORD Solutions Group Vice President Somesh Mukherjee said improved cost-efficiency, inference quality, and security standards are accelerating the rollout of generative AI tools across insurance workflows.
As the quality, cost of inferencing and security guardrails have matured, I’m seeing more and more applications starting to use gen AI
Somesh Mukherjee, ACORD Solutions Group Vice President
Cyber insurtech firm Coalition uses AI to streamline the underwriting process. Its system enables staff to process large volumes of data quickly, allowing underwriters, claims adjusters, and customer service teams to focus on higher-value work.
According to Coalition’s Chief Revenue Officer Shawn Ram, AI enables staff to shift from repetitive tasks to creative and analytical roles that drive innovation.
Using AI Agents in Insurance
One of the new AI applications involves using AI agents to help brokers with a multitude of tasks.
Flow Specialty said in a statement that its platformless AI model can “take meaningful work off brokers’ plates.”
These AI agents are trained on insurance materials such as specimens, quotes and carrier risk appetites, as well as the ongoing transaction of communication material and critical decision points.
The models are able to cross-reference materials, learn from broker interactions and evolve continuously, Flow said. AI agents can take on tasks such as risk appetite matching, generating quote comparisons and constructing review materials.
Flow said its human brokers are always in the loop, revising and overseeing the output of this workflow engine.
Flow Specialty said its technology is designed to help small and medium business segments that are often overlooked in favor of larger enterprises that yield larger commission dollars.
Flow Specialty said it used agentic AI to develop these AI agents. Agentic AI software is a type of artificial intelligence that uses large language models, machine learning and natural language processing to make decisions and perform tasks independently, according to IBM.
The software can help make decisions and perform tasks without human intervention, according to Salesforce Inc. These systems are able to learn from their interactions and adapt to new situations, improving their performance over time (see AI Promises to Revolutionize P&C Insurance).
Kristen Bessette, who was recently hired as chief data officer at Zurich North America, also spoke about the value of agentic AI, comparing its use to everyone getting an intern to help them do their job. She said those using it can “interact with data naturally, asking questions that pop into their head as they look at data and information.”
AI Enhances Insurance Claims and Underwriting Efficiency
Underwriters and claims adjusters manage large volumes of data to assess risk and evaluate claims. Artificial intelligence is being used to process, summarize, and present this information faster and in a more digestible format.
Generative AI improves access to relevant data across documents and systems, allowing professionals to focus more on analysis and decision-making.
Anthony Habayeb, CEO of Monitaur, explained that AI, as a modeling tool, aligns with the insurance sector’s foundation in modeling systems.
Monitaur supports insurers in managing ethical governance across these systems. Habayeb emphasized that AI allows underwriters and adjusters to condense information from diverse sources efficiently (see Increasing Use of AI Technology Triggers Insurance Claims).
That level of sophistication to route claims to the right place at the right time, rather than use the first-in, first-out methodology, can be an effective use of AI.
Shawn Ram, Coalition
At Liberty Mutual, AI copilots function as virtual assistants. According to Global CIO Monica Caldas, 25% of the company’s employees use LibertyGPT to streamline tasks, improve productivity, and save time.
The tool extracts information from multiple document types and delivers it to underwriters in a consistent format, supporting faster risk selection and pricing decisions.
Liberty Mutual also applies AI to compare documents, such as policy endorsements and regulatory updates.
For example, its crop insurance unit, RCIS, uses AI to detect annual procedural changes in federal documentation, saving analysts significant review time and minimizing the risk of oversight.
Insurtech Turvi developed a generative AI tool focused on the early stages of claims processing. This application directs adjusters to specific policy clauses without the need for manual searches (see How Insurers Can Get Best Results from Artificial Intelligence Tools?).
CEO Kenneth Tolson stated that the engine provides exact citations from policy documents, improving speed and precision. It also evaluates factors like coverage limits, sublimits, and coinsurance, assisting adjusters in making informed decisions.
Beyond internal efficiency, the tool also aids customer education
Tolson noted that one client broker wanted to share the system with end customers, highlighting its potential to improve coverage understanding.
For instance, a user could query the system and learn that flood coverage is missing, prompting them to consider additional protection.
Shawn Ram of Coalition pointed to AI’s potential for automated claim routing. Given Coalition’s global cyber portfolio with tens of thousands of policyholders, AI could identify claim types and match them with handlers based on expertise and availability, eliminating the inefficiencies of a first-in, first-out approach.
AI Improves Customer Communication
Cybersecurity remains a high-priority concern for most organizations, but translating technical risk information into accessible language remains a challenge. Shawn Ram of Coalition explained that most insurance buyers are not security experts, so AI helps bridge this gap.
Generative AI allows insurers to present complex issues in simpler terms, making it easier for decision-makers such as CFOs or risk managers to understand threats and identify solutions.
Ram added that AI helps classify companies across varying industries more accurately. By analyzing a company’s website and related third-party information, AI can suggest an appropriate industry classification using a range of predefined industry codes.
Coalition also uses AI to manage an estimated 400,000 to 500,000 broker service requests per year. These include submissions filed both manually and electronically.
The company applies a “copiloting” process, where AI converts incoming submission data into its system. Over time, performance data identifies areas of high accuracy and areas needing refinement.
This approach allows the company to concentrate human review only where necessary, improving efficiency.
Data Processing in Insurance Operations
ACORD, on the other hand, has applied AI to data extraction across insurance documentation. Its tool, ACORD Transcriber, uses intelligent document processing to extract information from more than 4,000 versions of over 800 industry-standard forms.
As data volume and computing power increased, the accuracy of the extraction model improved significantly.
Somesh Mukherjee, ACORD’s Vice President of Solution Architecture
The tool can now process both structured and unstructured documents across all insurance lines. For example, it can extract over 150 data points from the lengthy Market Reform Contracts used in Lloyd’s markets.
ACORD Transcriber identifies and returns key-value pairs with confidence scores, highlighting data relevant to each document type and customer case.
Mukherjee cited rental property insurance as a use case.
When a client submits a quote request with supporting documents, often scattered across emails and attachments, ACORD Transcriber can gather and organize relevant data for risk evaluation.
For structured documents such as ACORD Forms, the system avoids generative AI. For less structured content, generative AI is used, with clear notifications to users that some results may require human verification.
FAQ
Insurers apply AI to reduce workloads and support tasks like data processing, underwriting, and customer service. Companies such as Liberty Mutual, Zurich North America, and Coalition have implemented AI tools across claims, risk assessment, and internal operations to streamline decision-making and improve response times.
AI copilots act as virtual assistants to help employees complete tasks faster, while AI agents support brokers by handling functions such as quote comparisons and risk appetite matching. These tools enhance productivity by learning from previous interactions and refining outputs over time, particularly in small and medium business segments.
Coalition uses generative AI to translate complex cybersecurity information into simplified messaging for clients, enabling risk managers and CFOs to understand coverage gaps and risk solutions more clearly. The technology also assists in classifying industry sectors and improving submission handling through automation.
AI condenses data from multiple sources, delivering concise summaries to underwriters and adjusters. Tools developed by companies like Turvi and Liberty Mutual help identify relevant policy clauses, compare document versions, and evaluate coverage limits—cutting review times and increasing accuracy.
Agentic AI systems use large language models, machine learning, and natural language processing to perform tasks and make decisions independently. Flow Specialty and Zurich North America use such systems to manage document workflows and allow staff to interact with data more naturally.
ACORD Transcriber extracts and digitizes data from structured and unstructured insurance documents. It processes over 4,000 versions of 800+ ACORD forms and returns key-value data with confidence scores, supporting risk evaluations and quoting processes across all lines of insurance.
Coalition applies AI to assess incoming claims and assign them to handlers based on expertise and capacity, replacing the first-in, first-out model. This method improves claim resolution times and ensures that each case is handled by the most qualified team member.
……………..
AUTHORS: Edin Imsirovic – Director at A.M. Best Company, Anthony Bellano – associate editor, Best’s Review
