The banking, financial services and insurance sector (BFSI) has a problem. While nearly every industry leader agrees that digital transformation is business-critical, new research from Iron Mountain and HFS Research uncovers a stark disconnect between aspiration and action, with 78% of BFSI executives globally warning that failing to digitize could result in permanent competitive irrelevance.

The Back Office: No Longer Just Support

The back office has evolved from a support function to the backbone of operational resilience, regulatory compliance and differentiated customer service. Despite this, most organizations are struggling to move beyond legacy, paper-driven processes. While 81% of BFSI executives globally believe artificial intelligence will soon handle the vast majority of routine back-office tasks, only 13% have deployed AI at any meaningful scale. And while 77% believe the traditional back office will disappear within three years, only 21% are "radical transformers"—the organizations making bold moves to get there.

This leaves the BFSI industry to face an uncomfortable truth: The opportunity to achieve compliance, resilience and efficiency through digital and AI-powered operations is within reach, but only for those willing to move beyond incremental change. In today's BFSI sector, transforming the back office isn't just a lofty goal, it's become a business necessity. While many leaders are making ambitious plans to overhaul their core operations, turning that vision into reality remains a challenge.

The following seven hard truths highlight the disconnect between digital aspirations and the persistent realities of legacy systems, underscoring the struggle between commitment and true readiness for change.

Seven Hard Truths Reshaping the BFSI Back Office

1. Failing to digitize means falling behind permanently: Digitization has become a critical business success factor, with the overwhelming majority of leaders recognizing that organizations that do not act now could fade into irrelevance. Back-office transformation has shifted from a technological luxury to a strategic necessity.
2. Ambition outpaces action: While many organizations have expressed commitment to digitization and AI integration, only a small minority have successfully implemented AI-powered tools and systems on a large scale.
3. AI readiness is a workforce challenge: AI is poised to handle the majority of routine back-office work, with 81% of executives expecting AI agents to manage at least 75% of these tasks. Yet only 27% of organizations feel their teams are ready for this shift, representing a critical skills gap that could impede digital transformation efforts.
4. The "zero office" is still aspirational: While 77% of leaders believe the traditional back office will vanish within three years, replaced by an automated "zero office," only 21% are taking the bold steps needed to realize that vision. Most are hesitant, caught between legacy and opportunity.
5. Compliance drives change, but capabilities lag: Regulatory compliance is a top driver for back-office transformation. However, only 31% of BFSI firms have predictive, real-time compliance capabilities. As regulations accelerate, organizations must move from reactive to proactive compliance.
6. Investment and expectation are both high: BFSI firms plan to invest an average of $25 million each in back-office transformation over the next two years, with most demanding a return on investment in less than 24 months. This urgency requires clear priorities and a willingness to break from business as usual.
7. Only radical transformation yields true results: The research also spotlights a group of radical transformers—just 21% of respondents—who are investing in enterprise-wide reinvention. These organizations are already reporting higher revenue growth. In contrast, 79% remain stuck in incremental or limited transformation, risking long-term stagnation.

Radical transformers stand out not just in confidence but in results. These organizations treat back-office transformation as a growth engine, not a cost-cutting exercise, and they are seeing stronger revenue growth than their peers as a result. Their intent is matched by action: They invest more aggressively and target transformation that delivers enterprise-wide impact, not just incremental improvements.

For these leaders, customer experience is the guiding star. They view the back office as a direct driver of competitive differentiation, ensuring that every process ultimately supports better digital interfaces and client outcomes. Radical transformers also move quickly to adopt emerging technologies, building fluency in AI, automation and compliance tools that others are still piloting.

Crucially, they recognize that people are at the heart of transformation. Investment in upskilling and workforce readiness is central to their strategy, enabling teams to thrive in AI-driven, prompt-led environments. And while others focus on reducing risks or costs, radical transformers measure success by growth-oriented metrics, including innovation, customer experience and new value creation.

From Incrementalism to Bold Action: The Path Forward

The message from the research is unmistakable: Incremental change is no longer enough. Piecemeal digitization will not bridge the digital ambition gap or deliver the resilience, compliance and customer focus the industry demands. Only bold, enterprise-wide actions such as rethinking processes, investing in talent and scaling AI will separate tomorrow's leaders from the rest.

With BFSI firms accelerating investment and demanding rapid ROI, there is no room for hesitation. Those who move decisively now will shape the future of the industry. Those who delay will be left behind.

Businesses today are drowning in data. With everyone working from anywhere, the cloud taking over, and "bring your own device" (BYOD) policies running rampant, data is spread out, moving quickly, and constantly changing.

In the meantime, cyberattacks are getting smarter and faster. Collaboration tools have made sharing data extremely easy, and even GenAI is leaking data. As if that's not enough, compliance requirements are constantly evolving. With limited budgets, understaffing, and expansive skills gaps, it's no wonder IT and security teams feel overwhelmed.

Operationalizing data security has been a challenge for decades. Despite costly investments and countless hours of labor, admins are still flying blind. Legacy data security tools that require regex, trainable classifiers, or other pattern-based methods catch only a small fraction of sensitive data and bury IT teams in false positives.

The good news is that there are new, modern data security governance platforms available today that have ditched the legacy approach. In particular, businesses should seek solutions that leverage context-aware AI for discovery, risk monitoring, and remediation that can deliver the following benefits:

Superior visibility into their data: To effectively protect sensitive data, organizations first need to know precisely what data they have, where it's hiding, who's peeking at it, and how it's being shared.

Context-aware AI scans each data record in its entirety and can not only locate personally identifiable information (PII) and payment card information (PCI) but can even find things like intellectual property (IP) and other critical business records that other tools miss since this data doesn't usually contain patterns. Additionally, AI can identify duplicate or near-duplicate data, as well as the category and subcategory of each record. For example, it knows the difference between a bank statement and a corporate tax form or a resume versus a job application. Having this level of granularity enables security teams to make better-informed decisions when assigning classification labels, establishing where data should be located, or setting access and retention policies.

Stopped sensitive data leakage: Not only must security teams make sure that employees and third-party contractors aren't accessing data that they shouldn't, but they also must ensure that the users who are authorized to access it aren't sharing it. They should seek a solution that helps them contextually discover, monitor, and protect their sensitive data, not just at rest, but also as it travels to ensure that it isn't being shared with unauthorized users, personal email addresses, file sharing applications, social media, or GenAI applications.

Enabled GenAI without expanding the attack surface: GenAI is reshaping our world in real time. Tools like Microsoft Copilot, ChatGPT, Perplexity, and Google Gemini are changing how we make decisions, solve challenges, create content, and engage with others at work and home. But while they bring greater operational efficiencies, improved decision-making, and reduced costs, they also introduce significant data security risks.

Organizations need a solution that helps them identify when employees are using unsanctioned, or "shadow" GenAI, so they can regain control and keep their data secure. They also need to ensure that, regardless of where their data is located, it is accessed by the correct identities, at the appropriate times, and for the intended purposes. A truly comprehensive data security governance solution will enable them to define guardrails on what type of data should be blocked or redacted by groups and for each GenAI application and help them curate data when training their own proprietary GenAI workloads.

Aced regulatory compliance audits: Regulatory frameworks help businesses mitigate risks, operationalize processes, and maintain customer trust. But mapping security controls to these frameworks can quickly feel overwhelming. Adding further complexity, different industries and regions can have widely varying data handling and classification requirements. Businesses need a clear view of their compliance status, tools to fix issues, and peace of mind that they're not one audit from disaster. They should seek a solution that provides a dashboard showing their current compliance status with all relevant regulations and security controls, as well as support for custom frameworks. They also need granular visibility into all data records that violate compliance, with the ability to remediate them directly within the platform.

Maximized effectiveness of existing security tools: Tools like zero trust network access (ZTNA) and cloud access security broker (CASB) don't scan data to decide whether to allow or block access. Instead, they enforce policies based on labels, so if those labels are wrong or missing, they could either leak sensitive information to unauthorized users or block access needed for productivity. Context-aware AI and autonomous classification help ensure that sensitive data is labeled correctly and only accessible by authorized individuals.

Faster ROI, smarter policies, and less stress: Context-aware AI significantly speeds up the data discovery process and saves countless hours that administrators used to spend on algorithm tuning and chasing false positives. However, since new data is constantly generated and is always changing, capturing only a snapshot of the data at a single point in time is not enough. Security teams can save time and enhance data protection by implementing a solution that continuously monitors data, flags risks, and automates remediation steps. Choosing a provider that offers managed services can also reduce the burden on overstretched security teams by providing data security experts to assist with tasks ranging from deployment to training their teams on the platform, building a data governance roadmap, mapping classification labels, reporting, and tracking continuing progress toward their objectives.

AI agents aren't just another layer of automation—they mark a fundamental shift in how insurers can scale decision-making and operations. Unlike traditional tools, they can interpret context, provide recommendations and carry out tasks across multiple systems. For insurers, this isn't about just answering questions—it means executing real work and driving measurable outcomes.

For example, an underwriter can ask an AI agent to review broker submissions, extract risk data and suggest pricing tiers based on historical patterns. A business analyst can use an AI agent to analyze customer lifetime value and identify new retention strategies. A product manager can even have an AI agent configure new insurance products based on specific business requirements. These agents accelerate operations and improve efficiency while leaving judgment and final decisions in human hands.

AI Experimentation Is Not Enough

The potential is clear, but the reality is more complicated. According to Boston Consulting Group, 67% of insurers have experimented with AI, but only 7% have scaled it across their organizations. That means the vast majority remain in pilot mode, running isolated experiments that rarely expand into enterprise-wide capabilities.

That gap between promise and practice is where insurers risk falling behind. AI agents can deliver real value, but not if they remain trapped in proofs of concept. Scaling requires more than one-off pilots—it demands modern infrastructure, aligned leadership and interoperable systems that can evolve alongside the technology itself.

AI Agents Require Modern Infrastructure and Interoperability

Several technological obstacles keep insurers from deploying AI agents at scale. Legacy systems still dominate many organizations, making it difficult to connect AI to core functions like policy administration, billing and claims. Data is often fragmented, inconsistent and locked in silos, limiting the usefulness of even the most advanced models.

Even when insurers modernize their infrastructure, interoperability quickly becomes the new barrier. Today's AI ecosystem is highly fragmented, with each platform requiring custom development to connect with insurance workflows. The result is a patchwork of brittle integrations that are expensive to maintain and risky under real-world demands. Technical debt and compliance pressures only add to the complexity.

This creates vendor lock-in. Carriers often stay with a platform not because it's the best fit but because switching would mean rebuilding their entire AI infrastructure from scratch. The consequences are serious: Innovation slows, costs rise and insurers lose access to emerging capabilities that could deliver better results.

Enter the Model Context Protocol (MCP)

There have been several attempts to solve the AI interoperability challenge. Solutions like LangChain provided some help but locked organizations into specific frameworks, while function calling still required custom glue code for each connection. These early frameworks proved the demand for better connectivity but also exposed the limits of proprietary approaches.

In contrast, the MCP, introduced in 2024 by Anthropic, establishes a universal, open protocol—similar to USB for hardware—that lets organizations write a connector once and use it across different AI models and providers. This standardization eliminates redundant work, enables clean separation between data sources and AI applications, and creates a true plug-and-play ecosystem for AI agent connectivity.

For the insurance industry, the implications are significant. MCP allows AI agents to execute workflows securely, with auditability and governance built in. It enables portability, so organizations can switch AI providers without rebuilding their integration layer. And it accelerates innovation, since new AI tools can be adopted faster and with less friction.

MCP isn't perfect, but it's the most widely adopted solution so far—and a major leap forward in enabling open, interoperable AI systems. That's why it has quickly gained traction among enterprise software leaders including Salesforce, Snowflake, Atlassian, Hubspot and many more.

How Core Platforms Can Deliver AI Connectivity

MCP solves the interoperability challenge, but it does not address the underlying data problem. AI agents are only as effective as the data they can access. If insurers rely on legacy systems with siloed or inconsistent data, even the most advanced AI deployments will underperform.

This is why insurers need modern core platforms built for data fluency—the ability to access accurate and complete data whenever it's needed, in whatever form the business requires. A data-fluent core platform provides:

• Cloud-native data availability and performance to support real-time workflows
• Flexible data access, such as open APIs, data lakes and event streams
• Complete and governed data with metadata, lineage and auditability for compliance

When paired with MCP, a data-fluent core creates the ideal foundation for agentic AI. It ensures that AI agents can connect seamlessly to critical workflows while reasoning across high-quality data. Together, they unlock not just isolated efficiency gains but the potential for enterprise-wide transformation.

The Path Forward

Even with interoperable systems and data-fluent cores, AI agents cannot operate in a vacuum. In a regulated industry like insurance, transparency and accountability remain nonnegotiable. Human-in-the-loop governance—reviewing recommendations, validating outputs, and ensuring fairness—will be essential to earning trust and meeting regulatory requirements.

The insurance industry has reached an inflection point. AI agents are powerful enough to reshape underwriting, analytics, and product design, but scaling them requires both standards like MCP and modern core systems designed for AI connectivity. By embracing open, interoperable architectures, insurers can avoid vendor lock-in, reduce complexity and accelerate innovation.

The winners will be those who understand that AI is not just about smarter models—it's about building the infrastructure that allows those models to thrive. With the right foundation in place, insurers can finally move beyond pilots and unlock AI as a true engine of innovation and growth.

Congress's sweeping reductions to Medicare and Medicaid funding have set the stage for a decade of disruption in the U.S. healthcare system. The headline numbers are staggering: nearly $1.4 trillion in combined cuts over 10 years, reshaping the safety net programs that tens of millions of Americans rely on.

For insurers, the issue goes deeper than dollar amounts. It's about communication. Every coverage adjustment, every eligibility requirement, and every treatment approval must be explained and confirmed, often across multiple parties, before care can proceed. CIOs are suddenly staring down a future where communication infrastructure is the backbone of the entire business, rather than just IT.

Imagine a patient awaiting a time-sensitive surgery, only to have it postponed when a pre-authorization notice is delayed or lost in the system. Follow-up care could be delayed if the insurer never receives a discharge summary. Even something as routine as a coverage update arriving late can cause panic and confusion for a family already under stress.

Rural providers face even more formidable challenges. Small hospitals and clinics are already battling staff shortages and financial strain, and with $137 billion in cuts looming, those pressures are expected to intensify. Many of these facilities still rely heavily on fax because inconsistent broadband access in rural America and limited budgets for digital transformation mean older technologies remain a lifeline. As more patients move from public to private coverage, insurers must prepare for a surge in fax-based communication, not a decline.

This dual reality of modern digital channels and legacy tools means insurers need communication strategies that bridge both worlds. CIOs can't afford to let legacy systems create bottlenecks, nor can they risk patient trust by relying on generic digital tools.

One technology stands out: cloud fax. Unlike traditional fax servers, cloud-based faxing is scalable, transparent, and compliant with industry standards like HIPAA. It integrates smoothly with on-premise, hybrid, and cloud environments, reducing ineffective workflows and ensuring sensitive documents move quickly and securely between providers and insurers. Costs decrease, visibility improves, and compliance boxes get checked without slowing operations.

So where should insurers start? A practical roadmap for CIOs includes three steps:

1. Assess. Identify the customer segments and workflows most at risk. How many policyholders will be affected? Which communication channels will see the heaviest surges?
2. Evaluate. Test your current systems under pressure. Can they scale? Do they deliver consistently? Are compliance and redundancy baked in?
3. Act. Move decisively toward modern, cloud-native platforms that can flex with demand. Partner with providers that understand the stakes in healthcare — not just IT vendors but specialists in secure, regulated communication.

It's tempting to view these challenges as purely technical. But at their core, they're about people. Patients who don't know whether they're covered, doctors waiting for a green light before treating someone in pain, or families making decisions under enormous stress. Every communication failure ripples outward into real lives.

That's why, in the post-cuts era, insurers must treat every message as a lifeline. Those who invest in resilient, secure communication workflows today will be the organizations patients and providers trust tomorrow. Those who hesitate risk finding themselves overwhelmed at best, irrelevant at worst, in a healthcare landscape that isn't slowing down for anyone.

Generative AI (GenAI) promised insurers quick wins, yet most pilot programs stall out before they can deliver real business value. In fact, a recent MIT study found that nearly 95% of generative AI initiatives never move past the initial pilot stage. Datos Insights, an insurance research firm, concurred.

This isn't just a missed opportunity, though. It's a red flag. With AI moving faster than any technology that insurers have adopted before, failing isn't an option. Companies that can successfully adopt AI are those that learn to fail fast, iterate quickly, and focus on practical applications.

Four Common Pitfalls Holding AI Back

If AI is so powerful, why are nearly all pilots failing? For insurers, it often comes down to making the same missteps over and over again. These aren't technical failures as much as strategic ones. Below are the four most common mistakes insurers make when rolling out AI — and how to avoid repeating them.

Mistake 1: Confusing Building for Innovation

Many insurers believe success requires building AI capabilities in-house. Yet MIT's research shows vendor-built solutions succeed twice as often as internal builds.

That doesn't mean "buy and forget." Startups carry vendor risk, and custom development is rarely worth the cost or maintenance. The smarter path is to work with enterprise platforms that insurers already use. Providers like Microsoft, Salesforce, and Amazon continue to expand their AI services, offering reliable, secure, and scalable options without requiring teams to reinvent the wheel.

Mistake 2: Chasing the Shiny Object Instead of the Sure Bet

Too often, AI budgets flow into customer-facing applications such as chatbots, lead-scoring tools, or digital assistants. These may look impressive in a board deck but are difficult to validate and introduce risks insurers aren't prepared to manage.

The fastest return on investment (ROI) is usually in the back office. Automating document ingestion, workflow routing, and data extraction saves thousands of hours and frees staff to focus on higher-value work. These high-volume, repetitive processes are exactly where AI performs best, yet they're often ignored in favor of harder-to-execute, more glamorous projects.

Mistake 3: Expecting Perfection Out of the Gate

We often judge AI by unrealistic standards. We'll allow a junior underwriter a learning curve, but if AI makes mistakes on day 1, it's branded a failure. Like a new employee, AI improves with use, with accuracy and efficiency increasing over time.

Workforce anxiety compounds this challenge. If employees fear AI could replace them, they're quick to dismiss its early missteps. Leaders need to reframe the narrative: AI isn't about replacing jobs but about removing repetitive tasks so people can focus on higher-value decisions. Success depends on setting reasonable expectations and building trust in the process.

Mistake 4: Overthinking Instead of Taking Action

Lengthy RFPs, demos, and vendor evaluations can consume months or even years, creating the illusion of progress while manual processes remain unchanged.

Ironically, many insurers already own the AI tools they need through enterprise licenses. Capabilities like Azure Document Intelligence, Power Automate, and Copilot services can automate document intake, claims routing, and workflow support right now. The fastest path to value is often activating existing capabilities rather than prolonging procurement cycles.

Taken together, these missteps explain why so many pilots stall out — but they also highlight the path forward. The insurers that are breaking through have a very different playbook.

What Successful AI Programs Do Differently

Not every insurer is struggling. A small percentage — 5%, in fact — are moving beyond pilots and seeing measurable results. Here are some common traits they share:

• They narrow the focus. Instead of chasing enterprise-wide transformation, successful organizations zero in on a single pain point. Solving one operational problem creates quick wins, builds credibility, and sets the stage for expansion.
• They start with the obvious, not the flashy. Rather than automating complex underwriting decisions, they tackle the "boring stuff" first, such as email routing or workflow handoffs. These repetitive tasks are high volume and easy to supervise. And while they deliver clear ROI, they also provide a non-threatening way to introduce AI to the organization.
• They execute quickly. Long planning cycles kill momentum. Successful programs prioritize speed by testing, validating, and deploying in weeks. Short feedback loops enable them to refine models in real-time and maintain value flow.
• They partner with proven providers. Rather than betting on untested startups or building everything internally, they lean on the AI services offered by large, established cloud vendors. This reduces risk, simplifies integration, and ensures security and compliance standards are met.
• They set realistic expectations. AI doesn't need to be flawless to be transformative. If it outperforms the manual process, then it's a win. Successful insurers measure against that baseline, not against perfection.

The bottom line: These organizations succeed because they've redefined what success looks like. They don't expect AI to reinvent the business overnight. Instead, they use it to quietly, steadily strip out inefficiency, building both measurable ROI and organizational trust in the process.

Creating a Way Forward

Insurance leaders have a choice: keep overengineering and overpromising, or simplify, act, and deliver. The technology is ready, the platforms exist, and the use cases are obvious.

What's missing is the discipline to execute with focus and speed. Insurers that figure this out in 2025 will gain compounding advantages in efficiency, cost savings, and customer satisfaction. Those that don't will still be explaining to boards why their AI initiatives haven't moved the business metrics that matter.

Don't be part of the 95%. Join the 5% who are making the right choices.

The global insurance industry is built on the concept of helping people at a time of great need. This extends beyond assisting policyholders by helping to protect their homes, businesses and so much more to include becoming a force for good across our communities through charitable giving and volunteerism. Every day, professionals across our industry roll up their sleeves to offer their time and talent, and to give generously, in support of the communities where they live and work.

Our industry provided $1.3 billion in charitable contributions in 2023, along with more than 500,000 professionals giving of their time to volunteer, according to the most recent data gathered at the Insurance Industry Charitable Foundation. These figures reflect 100 insurance and insurance-related organizations.

The insurance industry also recognizes that helping others is not only good for the community but good for business. Positioning our board member companies at the forefront of community involvement and highlighting their social impact programs can help showcase the good the industry does while focusing attention on those in need.

Ours is an industry that appreciates the power of collective strength in working together to make a greater impact. At the heart of many influential efforts is our organization – The Insurance Industry Charitable Foundation (IICF), which, by working with the global insurance industry for more than 30 years, provides grants, volunteer service and leadership programs throughout the U.S., U.K. and, beginning this year, in Canada.

Each October, IICF hosts a global Month of Giving, celebrating insurance industry volunteerism throughout the year and highlighting philanthropic commitment in action. In this article, we'll talk about some insurance industry initiatives that make a difference in local communities and provide best practices to build a successful philanthropic program.

The New Normal

The evolution of work over recent years has had a significant impact on philanthropy and charitable giving. Whether operating in a remote or hybrid environment, a daily 9-to-5 or in-the-office structure is no longer a reality for many insurance professionals. As such, the ability to find and connect with people is critical, along with turnkey avenues for philanthropy given that people are spread across multiple locations, geographies and time zones.

IICF's Fill the Truck Food Drive is an example of a turnkey initiative that meets people where they are. Created and implemented during the pandemic as a safe, socially distant way to donate much-needed food, Fill the Truck also provides a pathway for various areas of connection in the form of in-kind donations and financial contributions from an organization to individuals donating and facilitating the collection and delivery of food. Since its inception, this program has grown to deliver thousands of meals across the West and Southeast U.S. regions through the IICF's food bank partners.

Another distinct challenge when navigating philanthropy in today's business environment is striking the right balance between corporately supported initiatives and causes and the varied, and often more locally based, charitable passions of individual employees. Both are important and carry a considerable role in shaping corporate culture, and the quality of business, employee and community connection. Embracing both a top-down and bottom-up approach is important, helping to ensure that the philanthropic avenues are not only strategic, but also authentic in their connections.

For instance, one long-serving IICF board company, Brown & Brown, had been seeking to engage its summer interns with philanthropy in a meaningful way. Through the organization's Next Gen Connect Program, IICF helped facilitate connections with nine nonprofits from around the country. The 61 Brown & Brown interns developed social media campaigns, created impact videos and supported several functions that small and medium-sized nonprofits do not have the capacity to coordinate. The project instilled a sense of purpose among the group of interns and goodwill toward our industry, while providing fulfilling and relevant experiences.

This is just one of thousands of volunteer efforts from our tremendous partners and their colleagues. Each year IICF features the contributions of our Key Partner Companies, those companies supporting us at the highest leadership level, in an annual publication that shines a spotlight on their extraordinary contributions through charitable giving, volunteerism and innovative industry leadership. The impact can be viewed in the 2024 IICF Insurance Industry Philanthropic Showcase.

Setting the Foundation

I have always believed the building blocks of a successful community outreach program already exist within each organization; the key is properly identifying this foundation through the organization's collective values. For example, a company's clients – and their own employees – are already connected within their own communities and are acutely aware of their neighbors' needs. To that end, insurance leaders should aim to gain a deeper knowledge of programs already in place and expand on them to benefit the community.

Second, I like to urge organizations to benchmark participation and effectiveness of any programs in place, the same as any successful business would benchmark operational goals. Good measurement of outcomes will enable a company to understand its current position and determine how to effectively develop or expand its philanthropic strategy.

Finally, companies should identify the top performing champions of philanthropic programs and empower them to drive and grow those efforts. These "doers," are the champions who are truly passionate about connecting the community and the selfless purpose behind charitable giving. Organizations that can effectively harness that employee passion will reap the rewards – including a positive reputation across the community, and enhanced employee recruitment and retention. Some thoughts for consideration to develop a greater employee connection:

• Cultivate connections. Effectively communicate the reason for launching a charitable campaign or a coming volunteer event to drive employee engagement.
• Make involvement easy. Create a variety of options for engagement and opportunities to connect. Not everyone has the time, ability or financial resources to contribute to every single cause. Make it easy for people to engage by giving them opportunities to amplify a philanthropic message via social media, or throughout the community without needing to commit time or financial contributions.
• Share the successes. Contextualize the impact of a particular donation or volunteer effort and celebrate the successes. Make your teams aware of how their contributions have made an impact on a particular campaign, mission or nonprofit organization. This can be done through sharing success stories or clearly identifying how donated funds will be deployed in the community.

The IICF is privileged to have nearly 300 board companies, and more than 800 individual insurance professionals serving on its boards and committees across the U.S., Canada and the U.K. These include organizations, individuals and teams passionate about doing good in the world and making our communities better for all. Building even stronger connections among employers, employees and the community benefits all involved. And participation in our Month of Giving is a great start - find out how you can get involved by visiting https://www.iicf.org/.

For decades, insurers used a broad-brush approach. They categorized people based on limited and generalized information. But this old model doesn't help today, as no two individuals are alike. From driving habits to home security, everything is unique. How can insurance businesses accurately price risk when the world is changing faster than a spreadsheet can track?

Thankfully, AI-based underwriting software systems have got insurers covered. These are changing underwriting from a slow, manual process into a dynamic conversation, one that includes actual risk and ensures the system is fairer, faster, and more accurate for everyone.

To truly appreciate the power of AI, let's first understand the challenges of old-school underwriting processes. It is through these issues that we recognize what AI-based underwriting software offers insurers and how they can capitalize on this opportunity.

Why Are Traditional Insurance Underwriting Methods No Longer Effective?

For decades, underwriters have been the backbone of the industry, making careful judgments based on their expertise and available data. Though this approach served the purpose well for years, it presents hurdles in terms of accuracy and profits. Here are some of the major hurdles insurance underwriters face:

1. Data Silos and Manual Entry

Important information often remains trapped in separate departments, either in physical files or online folders. What's more frustrating is manually entering and reconciling this data. Besides consuming time, this manual process makes room for human error, which can lead to incorrect risk assessment from the very start.

2. Static Risk Models

Insurers have always relied on historical data, which has undoubtedly proved valuable. But by looking backward, insurers may miss out on emerging risks entirely. A model built on past weather patterns, for example, may not be the right option for assessing a property's risk in the face of today's changing climatic scenario.

3. Assessing Everyone Using the Same Lens

One-size-fits-all approach worked really well when data was simpler and sources were limited. The problem arises when insurers have to differentiate between individuals who appear the same on paper but have totally different risks in reality. This often leads to homogenized premiums, causing insurers to miss opportunities to attract and reward low-risk customers.

4. Slow Turnaround Times

In times when customers can get a loan or book international travel in minutes, waiting for days and weeks for the underwriting process feels unrealistic. Such long waiting times frustrate potential customers, putting insurers at a competitive disadvantage.

Given all these factors, it's clear that insurance underwriting needs a new approach. It should add to human expertise. Insurance underwriting automation software aptly serves the purpose.

How Does AI Affect Underwriting in Insurance?

AI makes processes intelligent, and underwriting in insurance is no exception. At its core, AI allows insurance underwriters to make choices that are backed by data. Let's see how:

I. Machine learning

This is the engine of AI. ML algorithms are fed vast amounts of historical data, such as millions of past applications, claims records, and outcomes. Instead of being explicitly programmed, these algorithms learn to spot the complex patterns and correlations that lead to a claim. They continuously improve their predictive accuracy as they process more data.

II. Predictive analytics

This is the primary output of ML. By understanding the patterns of the past, the software can predict the future probability of a claim for a new applicant. It answers the basic underwriting question, "What is the probability of a loss?" with a much higher degree of detail.

III. Natural language processing

A huge portion of valuable risk information is buried in unstructured text. This includes doctors' notes in a medical record, detailed descriptions in a claims report, and even regulatory filings. NLP allows the software to "read" and "understand" this text, extracting relevant facts and sentiments that would be impractical to evaluate manually. The best part? All this is done while adhering to strict privacy protocols.

IV. Alternative data

This is where AI truly expands the horizon of risk assessment in insurance. Advanced models go beyond traditional sources, such as credit scores and motor vehicle records, to also consider non-traditional data points. For auto insurance, this could be telematics data showing actual driving behavior. For property insurance, it could be IoT sensor data indicating the quality of a building's maintenance. This creates a much richer, more dynamic picture of risk.

Insurance underwriting platforms with all these capabilities have the power to turn underwriting from a static, form-based exercise into a dynamic, multi-dimensional analysis. This is equally beneficial for both insurers and insureds. Explore what more insurers can do with such advanced solutions in the next section.

How Does AI-Powered Software Improve Risk Assessment in Insurance?

The theoretical advantages of the latest underwriting software are undoubtedly compelling. But what convinces the stakeholders to actually adopt is measurable results. By using AI-powered underwriting software, insurers can speed up old processes and enhance the quality and fairness of risk assessment. Here's what they can do:

i. Personalized Risk Profiling

AI lets insurers shift from grouping people into risk buckets to evaluating each applicant as a unique individual. By synthesizing thousands of data points, the software creates an individualized risk score.

The result? Two 40-year-old non-smokers living in the same ZIP code can receive vastly different life insurance premiums. Why? Because the model considers one's fitness routine and regular health check-ups versus the other's inactive lifestyle. This fairness benefits both the insurer, which can price risk more accurately, and the customer, who pays a premium truly reflective of their individual situation.

ii. More Accurate Fraud Detection

The human eye is excellent, but it can miss subtle, complex patterns that might point toward fraud. AI algorithms excel at spotting anomalies and correlations that are invisible in a manual review. By assessing an application against millions of previous ones, the software flags inconsistencies. For example, a discrepancy between stated income and spending patterns obtained from alternative data, or a claims history that follows a suspicious pattern.

Thus, insurers can spot potentially fraudulent applications at the point of underwriting, preventing losses before a policy is even issued and protecting honest policyholders from bearing the cost of fraud.

iii. Higher Efficiency and Speed

One of the most immediate benefits is efficiency. AI-powered underwriting software solutions handle the routine, repetitive tasks that eat up a human underwriter's time. It can instantly validate data, run checks against external databases, and even make straight-through processing decisions on low-risk, standard applications.

This slashes turnaround times from days to minutes, meeting the need for instant gratification. Crucially, it also elevates the role of the human underwriter. Freed from mundane tasks, they can focus their expertise on complex, high-value cases that require nuanced judgment and empathy.

iv. Risk Insights

The latest underwriting software enables insurers to prevent risks instead of crying over them later. By connecting real-time external data feeds, such as climate models, geospatial imagery, and economic indicators, these solutions can tell what the future holds.

For instance, they can spot properties at increasing risk of wildfire due to changing vegetation density and drought conditions. Alternatively, they can also identify commercial properties in a supply chain that are more susceptible to specific geopolitical disruptions. This allows insurers to work with clients on risk mitigation strategies before a loss occurs, turning the insurer from a simple payer of claims into a genuine risk management partner.

Wrapping Up

The way insurers assess risk today is way different from how it was done, say, three to four years ago. It is no longer a craft meant only for experienced underwriters but has become a fair practice, one that ensures accuracy and transparency. And AI-powered underwriting software for insurance helps businesses take the jump.

Not long ago, managing money meant sitting down with a paper checkbook, a monthly bank statement, and a pencil. Each month, you'd receive a paper statement via post, manually tick off cleared payments and try to reconcile the numbers. This process was slow and error-prone and gave you little-to-no real-time visibility into your finances. It feels like a relic, but claims funds are still managed with the same outdated mechanics.

Claims funds remain scattered across stakeholders and systems. Banking structures are split between insurers and TPAs, while MGAs often sit in the middle, reliant on reports that arrive late or don't align. Across all parties, reporting lags behind reality. Reconciliation still depends on manual processes, and payments are often handled in batches rather than in real time. What should be a seamless financial flow instead resembles balancing a giant, industry-wide checkbook.

This fragmentation is more than just inefficient. It is a structural problem that drains liquidity, hides cash, slows payments, and consumes valuable time across the entire insurance value chain. Tens of millions of dollars often sit idle, reducing efficiency and yield. Accounts are routinely underfunded or overfunded. Teams spend hours reconciling balances when they could be focused on strategy and performance. The result is higher costs, lower margins, lost opportunities for investment, and a poorer customer experience at the very moment policyholders expect speed and certainty.

Hidden Costs Identified

The fragmentation of claims funds creates a series of hidden costs that ripple across the entire insurance ecosystem. While these costs may not appear on a balance sheet, their impact is felt in liquidity management, operational efficiency, profitability, and customer satisfaction.

1. Liquidity inefficiencies

When funds are spread across multiple accounts and stakeholders, balances are rarely optimized. Some accounts sit underfunded, limiting the ability to settle claims smoothly. Others hold excess cash that sits idle instead of being deployed or invested. For many insurers and TPAs, this can mean tens of millions of dollars not only tied up unproductively but, in some cases, effectively invisible, hidden across fragmented accounts with no clear line of sight.

This is not just a theoretical issue. A recent analysis of U.S. property and casualty insurers found that companies have increased their holdings in cash and other short-term liquid assets by more than 3% year over year, largely in response to rising catastrophe losses and unpredictable claim demands. While holding liquidity is prudent, the need to maintain such large buffers often reflects the limitations of fragmented systems and delayed reporting. Without clear visibility into cash positions, insurers lock up capital that could otherwise be put to more productive use.

2. Lost real-time visibility

Fragmentation obscures the true financial picture. With accounts scattered and reporting delayed, it is difficult for carriers, MGAs, and TPAs to know exactly where cash sits at any given moment. According to one recent industry commentary, insurers often struggle with "lack of balance visibility," especially when funds are handled across different accounts, in restricted currencies, or segmented by TPAs and insurers. This lack of real-time oversight hinders decision-making and increases the risk of errors, especially in times when claims surge and liquidity is needed most.

3. Operational inefficiencies

Reconciliation is still largely a manual process. Teams spend hours chasing reports, matching payments, and trying to balance accounts—much like the old days of reconciling a paper checkbook. According to the State of Claims Finance report, nearly eight in 10 insurers (79%) see these internal inefficiencies as a major barrier to timely claims payments, and 78% point to coordination breakdowns with brokers, TPAs, and banks as a source of friction. Only 1% describe collaboration between claims and finance teams as "highly effective," a telling sign that silos persist across the value chain. At the end of the day, this represents thousands of hours lost to administration, diverting talent away from analysis, strategy and improving performance.

4. Profitability impact

When money sits idle, balances are unclear, and teams spend hours on manual work, profits suffer. Insurers lose out on opportunities to put cash to better use, everyday costs remain high, and the gap between revenue and expenses narrows. Competitive pressures in the market make these inefficiencies even harder to absorb. Across U.S. insurance markets, rate dynamics are increasingly uneven. Some lines continue to see upward pressure, while others, such as property and certain financial lines, are beginning to soften as more capacity returns. Catastrophe exposure, such as the impact of California wildfires, adds another layer of volatility that puts further strain on margins. In this environment, inefficiencies in claims fund management are no longer a minor inconvenience. They have become a direct drag on profitability.

5. Customer retention costs

Perhaps the most damaging consequence lies in the policyholder experience. Delays in claims payments erode trust. For example, in the 2025 J.D. Power Small Commercial Insurance Study, customers rated "problem resolution" and "ease of doing business" as among the top drivers of overall satisfaction. Small commercial customers who experience poor problem resolution are much less likely to renew.

In an environment where even small mistakes or delays can push buyers to shop elsewhere, this becomes expensive. The cost to acquire a new policyholder far exceeds the cost to keep an existing one—with churn tied to claims experience being an accelerating risk, especially in specialty and smaller commercial lines where margins may already be thinner.

The Path Forward

If fragmentation is the problem, simplification is the solution. The industry faces a choice: continue investing in bespoke systems that are costly to build and slow to adapt, or embrace shared, modern infrastructure that enables real-time visibility, automated reconciliation, and faster payments. The goal is not to rip out everything that exists today, but to rewire the financial foundation of claims in a way that reduces friction across the value chain.

Progress will come as more stakeholders, carriers, MGAs, TPAs, and their partners, work from connected platforms rather than isolated silos. When that happens, cash is deployed more effectively, reconciliation becomes less resource-intensive, and policyholders receive funds with greater speed and certainty. The benefits go beyond efficiency. Faster, more transparent claims payments improve trust, and at scale can strengthen the resilience of the entire insurance ecosystem.

AI is transforming insurance, but most organizations aren’t prepared to take full advantage. CLARA’s Data Engineering as a Service (DEaaS) solves the data readiness challenge by converting fragmented inputs into clean, AI-ready intelligence.

Discover how leading insurers are:

• Unifying siloed data across systems and vendors
• Improving claims outcomes with agentic reasoning
• Future-proofing their AI investments with trusted data foundations

Learn how CLARA DEaaS can help your organization unlock the full potential of AI.

Read Now  

Sponsored by: CLARA Analytics

Recent reports on the improving dexterity of robotic hands have raised the prospect that humanoid robots will show up in big numbers in workplaces and homes within the next few years, with obvious implications for workers' comp and homeowners insurance. Many investors are all in on the idea: Securities analysts say some 75% of Tesla's $1.5 trillion of market value stems from optimism about its prospects for "embedded AI," in its cars and Optimus humanoid robots. Hypemeister-in-chief Elon Musk said this summer that the robots could generate $30 trillion (that's trillion, with a "t") in annual revenue for Tesla.

Color me skeptical. I think specialized robots, a la the roughly 1 million in use in Amazon warehouses, will continue to proliferate rapidly but believe it will be decades before humanoid robots can function like people in the home and workplace.

And it just so happens that a thorough takedown began making the rounds recently, explaining in a far-more-learned way than I could just why human-level dexterity remains so far off for our machines. 

I'll share. 

The bullish case for humanoid robots goes more or less like this article from Bain, which begins:

"Dexterous, bipedal robots with general intelligence are advancing faster than many expected, and they’re quickly becoming economically viable. Within five years, robots will likely be able to perform a wide range of physical tasks at a cost that rivals or beats human labor. Adoption is poised to accelerate across industries, from manufacturing to food service, healthcare, and even construction."

The piece adds that "robotic mobility and dexterity are reaching human levels" and that "cost parity is within reach" for robots and human labor.

Rodney Brooks, a professor emeritus of robotics at MIT, counters that, while "the general plan is that humanoid robots will be 'plug compatible' with humans and be able to step in and do the manual things that humans do at lower prices and just as well..., believing that this will happen any time within decades is pure fantasy thinking."

He zeroes in on the dexterity issue, where he says humanoid proponents are making a false analogy to other AI-based systems — image labeling, speech to text and large language models (LLMs) — that have made exponential progress. He says those three benefited from decades of research that provided detailed, digital descriptions of what constituted an image, speech and the text that LLMs have been trained on. That baseline needs to be there, Brooks says, before you can turn machine learning loose and get the kinds of near-magical improvements we've seen with images, speech/text and LLMs. Yet researchers and developers are, he says, simply showing their robots videos of people handling objects and counting on the AI to figure out how to copy the movements. 

Brooks adds that humanoid robots have nowhere near the sensitivity of a human hand, with its "about 17,000 low-threshold mechanoreceptors in the glabrous skin (where hair doesn’t grow) of the hand, with about 1,000 of them right at the tip of each finger." 

He says walking is the other main issue with humanoid robots. He acknowledges that he's seen robots about half the height of humans maneuver smoothly among people in somewhat chaotic environments but notes that robots don't (in fact, can't, he says) walk as smoothly as humans and says the issue becomes far more complicated if the robot is the full height of a human, as it needs to be to take over most human tasks. The tendency is to think that doubling the height merely doubles the complexity, but you're also doubling the width and the depth, so you're actually having to deal with roughly eight times the volume and mass.

I'll add the cost issue. Musk's claim that Tesla can generate $30 trillion a year in revenue from humanoid robots is based on a price of $30,000 apiece. I don't like doing laundry or putting away dishes any more than the next person, but I'm not going to spend $30,000 (plus some annual fee for maintenance) just to have a robot do the chores, then go stand creepily in a corner until I assign it another task. 

I'm not at all suggesting that robots won't play a huge role in our future — just that they will be in the workplace, not the home, and that the robots will be specialized, not humanoid. I think the future for manufacturers, retailers, fast-food restaurants and many others looks a lot like what Amazon is pioneering with the sort of array of task-specific robots described in this New York Times story. 

As this related article in the Times shows, the results will still be profound for insurers, because hundreds of thousands of workers won't be hired for the sorts of jobs that run relatively high risks of injury — and that's just at Amazon. The article says Amazon expects to sell twice as many products by 2033 but will need some 600,000 fewer workers in its warehouses because of robotic automation (while needing some thousands of robot maintenance workers who, Amazon notes, will earn higher pay). 

Just don't expect the robots to look like you or to run your home like Rosey the Robot did for the Jetsons.

Cheers,

Paul