ICMIF Supporting Member Microsoft provides an insight into how mutual insurers can use AI to tackle risk and boost financial resilience.
For generations, the value proposition in insurance has been defined by risk transfer: when losses occur, insurers help policyholders recover financially. That role remains essential. But major long-term shifts across the global insurance landscape are now forcing a reimagining of customer value, profitability and growth.
Property and casualty (P&C) insurers face growing challenges, including macro-economic factors and cyber-attacks, but none is bigger than climate risk. Catastrophic events are nothing new, of course; what has changed is the scale and frequency of weather-related losses and the operational strain that follows. Swiss Re estimates global insured losses from weather‑related natural catastrophes will exceed USD135 billion in 2024, marking the fifth consecutive year insured losses topped USD100 billion, and underscoring a structural escalation in climate‑related risk[1].
[1] Swiss Re, “Hurricanes, severe thunderstorms and floods drive insured losses above USD 100 billion for 5th consecutive year, says Swiss Re Institute,” December 2024
The article was authored by Mona Kothari-Chitalia, Managing Director, Industry Advisory, Microsoft.
Published May 2026
In response, many insurers are rethinking how to best deliver customer value, profitability, and growth. Mutual and cooperative insurers are under sustained pressure to balance financial strength with their purpose of providing protection in an environment marked by increasingly severe risks and closer regulatory scrutiny. It is a challenge that AI is well suited to answer, helping to expand the role of insurers from risk transfer providers to proactive risk partners.
Insurers and AI: early adoption and opportunity
A 2024 survey by the International Cooperative and Mutual Insurance Federation (ICMIF) found that 62% of respondents were already using AI, with a further 19% planning adoption within the next year. In practice, however, most deployments were commonly concentrated in specific functional areas, such as supporting underwriting, claims processing, and customer interactions. About 67% of insurers expect AI to become more central to their operations, even as many cite data quality and talent gaps as key challenges[2].
[2] International Cooperative and Mutual Insurance Federation, “Balancing AI innovation with member-driven values at mutual and cooperative insurers,” February 26, 2025
According to a recent study by BCG, only about 7% of insurers have successfully scaled initiatives beyond pilots, with most efforts remaining fragmented across functions. The opportunity now is to move from isolated use cases to AI embedded across end‑to‑end processes, extending to more automated, interconnected workflows and setting the stage for a shift toward risk prevention[3].
[3] BCG, “Insurance Leads in AI Adoption. Now It’s Time to Scale.” September 4, 2025
As an ICMIF Supporting Member, Microsoft is collaborating closely with ICMIF and its members to drive meaningful transformation within the insurance industry, focusing on advancing digital innovation, fostering trust, and delivering solutions that create lasting value for policyholders worldwide.
How AI helps improve efficiency, service, and relationship management
Prevention does not replace excellence in risk transfer. Forward-looking organisations pursue both. They modernise service and core operations across the customer engagement cycle, while investing in prediction and prevention-oriented capabilities that help reduce future risk and strengthen long term resilience.
One area where AI delivers important benefits is in enabling faster and more consistent client service, such as helping representatives locate and validate policy information faster. For example, NFU Mutual (UK) is using Copilot for Sales and Microsoft Dynamics 365 to unify customer data and capture interactions in real time. It provides a centralised “single source of truth” that helps employees reduce response times and deliver more informed, personalised interactions.
Likewise, a new AI-powered application at one long established carrier lets client experience representatives search across 1.3 terabytes of policy documents and receive highly relevant answers in four to five seconds, with reported accuracy of up to 95%. This helps free up time for more personalised, empathetic interactions.
AI can also help streamline First Notice of Loss by ingesting call transcriptions, images and videos, and by guiding representatives to capture the right information in the first conversation, which can accelerate remediation.
In claims review, AI turns static documentation into insights that inform action. For example, one major global insurance and risk management firm built an internal AI platform that summarises complex claims files in minutes rather than hours, helping adjusters move faster and apply those insights more effectively across claims and client workflows.
In cases of widespread impact, such as a storm that causes power outages that result in many food spoilage claims, AI can route low-complexity claims through specialised agents that can help validate coverage, correlate weather data, detect fraud, calculate payouts, and generate audit trails.
These innovations use document processing, contextual summarisation, natural language interface and workflow automation, all of which can be used to help improve other processes across core insurance capabilities, customer service, and relationship management.
How AI helps with prevention and protection
The impact of prevention‑led approaches, whether applied to customer risk or enterprise risk, is twofold: financial resilience and stronger trust. This positions insurers as partners that mitigate, not just transfer risk for their customers.
Prevention‑focused use cases extend well beyond field‑level interventions, such as property risk scoring or event‑readiness outreach. Increasingly, they will focus on proactively reducing risks before disruptions, security incidents, or service failures occur.
The value of this sort of internal prevention model can be seen in one global risk advisory firm’s approach to improving decision quality under strict governance. To support more consistent, data-driven analysis, the firm developed an Azure-based AI platform that securely connects enterprise data and enables governed collaboration in fast-moving situations. During a recent natural disaster, teams combined near real-time satellite imagery with proprietary data to generate timely insights, helping clients assess impacts and plan responses.
AI also enables a key shift from paying claims after events happen to helping customers reduce exposure before losses occur. For example, one carrier has deployed hundreds of AI capabilities on Microsoft Azure to interpret and translate unstructured inputs in the form of images, reports, and emails in multiple languages, into clear, consistent risk signals for underwriters. This can improve the accuracy and timeliness of risk assessments, which supports better informed underwriting decisions and helps customers anticipate and reduce potential exposures before losses occur.
Emerging external data sources help improve risk prevention
Many types of prevention depend on spotting and interpreting early signals, often from outside of core insurance systems.
Using Large Language Models (LLMs) and machine learning, insurers can integrate third-party signals with internal data to help create new ways to refine risk selection, pricing, event readiness, customer outreach, and more. Sources such as research, disclosures, regulatory filings, sensor data, and geospatial imagery can have immense impact, provided they are reliably accessible.
Initiatives from Microsoft Research and AI for Good highlight advances in third-party data that can significantly enrich the power of predictive solutions:
- First, Aurora is a foundation model of the atmosphere that produces fast, high-resolution forecasts, especially during extreme and fast-moving conditions. For insurers and reinsurers, that means more timely environmental intelligence to support underwriting, catastrophe modeling, claims surge planning, and reinsurance response.
- Second, SPARROW uses solar-powered devices with cameras, microphones, and sensors to detect meaningful changes on the ground and send near real-time insights to the cloud. For insurers, it shows how AI and sensor data can enable earlier risk detection, faster intervention, and reduced loss severity.
Earlier and more precise forecasting supports proactive risk alerts, giving customers and commercial clients time to take preventive actions (for example, securing property or adjusting operations) and allowing insurers to coordinate with risk engineers, brokers, and public authorities. The objective is straightforward: improve lead time and decision quality so that fewer losses become large losses.
Priorities for success with AI and risk prevention
For leaders, realising measurable value from AI across the business, including enhancing prevention, can happen in a matter of months or quarters. Microsoft’s view of industry patterns indicates that successful approaches often prioritise the following:
- Define a clear strategy and start with a small number of high‑value, extendable use cases aligned to core business priorities.
- Build strong data foundations and effective governance.
- Balance innovation with credibility and responsible adoption.
- Focus on business-led re-architecting of processes, change management, including talent skilling, operating model refinements, and accountability
- Establish and commit to stretch goals with active leadership and resourcing.
Insurers who employ this comprehensive approach and tailor AI to their unique business requirements can improve the most critical aspects of their operations. As importantly, they can enhance prevention as an important part of their future growth strategies.
