These 2026 cybersecurity predictions offer insights into the trends that will be front and center in the coming year, particularly around AI and data protection.

There is a great deal at stake. The business world is on the verge of making significant AI breakthroughs, but the technology has some serious security gaps that must first be addressed. At the same time, regulatory environments around the globe are increasing pressure on companies and AI providers to establish provable, trustworthy practices. And while practical quantum attacks may be several years away, the "harvest-now, decrypt-later" threat is creating urgency for organizations to act today.

All of this is driving a critical need for organizations to ensure end-to-end protection of data at rest, in transit, and, most importantly, in use – without which, companies expose themselves and their customers to potentially irreparable harm.

The predictions below are intended to help guide organizations in embracing AI-powered innovation while maintaining persistent protection of sensitive data and ensuring regulatory compliance across regions worldwide.

1. Confidential AI: A Business Imperative

In 2026, enterprises will integrate AI more deeply into core operations, moving beyond experimentation toward scaled, business-critical deployments. This expansion will expose the limits of today's security measures and accelerate demand for "Confidential AI" — systems designed with built-in privacy, encryption, and trust guarantees.

Much like the early days of the Web, when open protocols gave way to HTTPS and SSL, organizations will shift from simply using AI to securing the full AI lifecycle – from data ingestion to model training and inference. As breaches targeting AI models and systems increase, companies will adopt proactive protection strategies by embedding privacy, encryption, and integrity controls directly into their AI architectures.

As enterprises advance their AI capabilities, Confidential AI will emerge as the new standard – embedding privacy and protection into every layer of the AI lifecycle. Through continuous, end-to-end encryption and confidential computing, organizations can train and run models securely, even on sensitive data. In the year ahead, growing demand for zero-trust AI ecosystems will redefine the landscape, making security the hallmark of enterprise AI rather than an afterthought.

Predictions 2026:

• Organizations will move beyond protecting perimeters and threat detection to securing models, data, and inference chains.
• Confidential AI, powered by continuous encryption and secure enclaves, will define the next phase of AI security.
• AI security will evolve toward "Confidential AI," where encryption and privacy-preserving computation become essential for trusted enterprise deployment.

2. Compliance and the Rise of Sovereign AI

In 2026, as AI compliance takes center stage in the enterprise, we'll also see the rise of Sovereign AI – nationally governed AI ecosystems that affect global data flows. As nations tighten restrictions on how models are trained, hosted, and shared, companies will face growing pressure to demonstrate compliance across multiple jurisdictions simultaneously. This will require organizations not only to meet their country's privacy and data security standards but also to comply with foreign laws governing AI transparency, data residency, and model integrity.

The next wave of regulation will focus on "trustworthy" models – AI systems that can prove, through cryptographic means, that data remains secure and private throughout the entire lifecycle. Governments, telecom companies, and cloud providers will need to go beyond contractual promises to offer verifiable assurances that they cannot see or misuse a customer's data or model. Model theft, data exfiltration, and misuse of AI-as-a-Service will rise sharply, forcing providers to deliver cryptographic evidence of confidentiality to satisfy regulators and clients alike. In 2026, enterprises operating in multiple regions will recognize that compliance and security are inseparable, and that continuous encryption will become the cornerstone of regulatory trust in AI.

Predictions 2026:

• Sovereign AI will push multinational compliance beyond borders, demanding alignment with overlapping global standards.
• Model-as-a-service providers will face rising risks of theft, exfiltration, and regulatory scrutiny.
• "Trusted models" will require cryptographic proof of data confidentiality and model integrity.
• Technologies such as FHE and secure inferencing will underpin compliance verification frameworks.
• Cryptographic assurance will become the foundation of AI trust, compliance, and cross-border collaboration.

3. Quantum Resiliency and Latency-Optimized FHE

In 2026, quantum computing – both a looming technological breakthrough and cybersecurity threat – will remain a key concern among enterprise CISOs. The "harvest-now, decrypt-later" tactic, where adversaries stockpile encrypted data to decrypt once quantum hardware matures, will accelerate demand for latency-optimized, quantum-safe encryption. When the day comes, traditional standards like RSA and elliptic-curve cryptography will be rendered obsolete by quantum algorithms, leaving unprotected financial data, health records, and AI models vulnerable.

The next wave of security innovation will focus on quantum-resilient encryption at scale, capable of protecting data without slowing real-time AI workloads. Latency-optimized, production-grade fully homomorphic encryption (FHE) enables computations on encrypted data, safeguarding information throughout the AI lifecycle.

Predictions 2026:

• The "harvest-now, decrypt-later" threat makes post-quantum migration urgent today.
• Latency-optimized FHE enables quantum-resilient AI inference without compromising performance.
• Quantum-safe, latency-optimized FHE will be recognized as essential infrastructure for securing AI systems worldwide.

In the coming year, the benefits of AI will only be accessible when the technology is built upon a tapestry of privacy and security. The collective, urgent need for proactive protection, data sovereignty, and quantum resilience is driving Confidential AI to become a prerequisite for enterprise competition and trustworthiness. By building security into the very fabric of their AI, organizations can confidently protect their most valuable assets, innovate across borders, and gain the competitive trust necessary to thrive in the global digital economy.

As we take our first steps into 2026, many enterprise leaders are focused on a familiar concern: how to keep employees' digital skills current. Strategies center on reskilling programs, digital academies, and AI literacy initiatives. These efforts are well-intentioned, but they target the wrong problem. The issue is no longer that employees lack digital skills. It is that the usable life of those skills has collapsed.

Technology is evolving faster than organizations can formalize roles, update training curricula, or redefine career paths. By the time a skill is taught, validated, and embedded, it is often already outdated. In this environment, treating digital fluency as something employees can "catch up on" is a losing strategy. Heading deeper into this year, enterprises must stop thinking about digital fluency as a destination and start treating it as a continuous operational capability.

The skill gap is widening despite investment

There is no shortage of data underscoring the urgency. McKinsey research shows that companies with leading digital and AI capabilities outperform competitors by two to six times in total shareholder returns. Yet McKinsey also reports that nearly six in 10 workers will require significant retraining before 2030, and fewer than half of job candidates possess the high-demand digital skills employers list in job postings. These gaps persist even as learning budgets expand.

The disconnect is structural. Traditional upskilling assumes relative stability, predictable tools, and slowly evolving roles. That assumption no longer holds. AI, automation, and workflow technologies are reshaping jobs continuously, often faster than organizations can document the change. No centralized training function can keep pace with that level of volatility.

Digital fluency is no longer an IT issue

Another constraint holding organizations back is the belief that digital fluency primarily belongs inside IT or data teams. That distinction has eroded. Business leaders now oversee AI-driven decision systems, automated workflows, and technology-enabled products. According to McKinsey's "We're all techies now" analysis, executives increasingly need foundational understanding of cloud architecture, data governance, cybersecurity risk, and AI trade-offs simply to perform their roles effectively.

In insurance, this shift is unavoidable. Underwriting, claims, fraud detection, regulatory reporting, and customer engagement are all shaped by technology decisions. When only technical teams understand how these systems function, the organization becomes slower, more brittle, and more exposed to operational and regulatory risk. Digital fluency must extend across underwriting, claims, compliance, operations, and leadership if insurers expect to adapt at speed.

Training cannot keep up. Adaptability can.

Faced with accelerating change, many organizations respond by expanding course catalogs, launching academies, or mandating certifications. These efforts provide value, but they do not solve the core challenge. Skills decay faster than training cycles can replenish them. What matters more is whether employees know how to adapt as tools, workflows, and assumptions change.

This is where the conversation needs to shift. Continuous learning is not about more content. It is about redesigning work so learning happens inside execution. McKinsey has found that while 80% of tech leaders view upskilling as the most effective way to close skills gaps, only 28% plan to meaningfully increase investment in the next few years, in part because traditional programs struggle to show return. Learning that sits outside real work rarely scales. Learning embedded into workflows compounds.

The insurance workforce will be judged differently

As digital skills become more transient, performance expectations must change. Employees cannot be evaluated solely on mastery of specific tools. They must be assessed on how effectively they adapt, challenge outputs, collaborate across functions, and apply technology responsibly.

For insurers, this means valuing underwriters who understand how models behave rather than simply how to operate them. It means claims professionals who can work alongside automation while exercising judgment in ambiguous cases. It means leaders who can interrogate AI-driven outcomes, governance structures, and risk exposure without relying exclusively on technical intermediaries.

The real challenge for 2026

The organizations that succeed in 2026 will not be the ones with the longest skills lists or the most certifications. They will be the ones that redesign work so adaptation is unavoidable and learning is constant. Managers will be expected to coach in real time. Workflows will be designed to expose employees to change rather than shield them from it. Technology will teach through use, not through periodic retraining.

The uncomfortable reality is this: no enterprise can train its way out of rapid technological change. Skills will continue to expire. That is not a failure of people or programs. It is a structural condition of modern work.

The real question for leaders heading into 2026 is not how to preserve digital skills, but whether their organizations are built to function when those skills inevitably expire.

Insurance fraud in property and casualty insurance is no longer dominated by isolated bad actors slipping through the cracks. Increasingly, the most damaging fraud comes from coordinated networks of claimants, service providers, and intermediaries who understand carrier processes and exploit their fragmentation. These schemes rarely trigger obvious red flags at the individual claim level, which is precisely why they are so costly and so difficult to stop.

The Scope and Impact of Fraud in P&C Insurance

The cost of fraud in insurance is staggering. According to the Coalition Against Insurance Fraud (CAIF), total insurance fraud (across all lines) is estimated at $308.6 billion per year in the U.S. Within the P&C sector, fraud represents a significant portion of losses. Industry analyses consistently estimate that 10% of P&C claims are fraudulent or involve fraud-related padding.

A recent report on fraud detection solutions estimates that roughly $50 billion in P&C paid claims may be fraudulent each year. Meanwhile, the FBI (via NAIC-supported research) has historically pegged total non-health insurance fraud — including P&C — at $40 billion annually. Another survey found that many insurers believe fraud costs represent 5-10% of their claims volume; some even reported that up to 20% of their claims were subject to suspicious activity.

These losses hit insurers' bottom lines directly through inflated payouts, but the effects ripple outward:

•  Premium increases: Fraud raises loss costs, which then contribute to higher premiums for honest policyholders.
• Operational drag: Investigating suspected fraud takes time, manpower and specialized expertise.
• Reputational risk: Organized fraud rings damage trust in the industry and can create regulatory scrutiny.

Why Network Fraud Evades Traditional Detection

Network fraud (also called collusive or organized fraud) refers to schemes where multiple actors cooperate to submit fraudulent claims or inflate losses. Rather than isolated "opportunistic" fraud, these are coordinated efforts that may involve:

• Claimants staging accidents or exaggerating damage.
• Repair shops or body shops submitting inflated or fake invoices.
• Brokers or agents steering business toward fraud-friendly networks.
• Collusion between claimants and medical providers, attorneys or investigators.

Because these networks are interconnected, detection is especially challenging. Traditional rule-based fraud detection — checking individual red flags on a claim — may not catch the systemic patterns. That's why more insurers are turning to network analytics, which analyzes relationships across participants to spot suspicious clusters.

In academic research, for instance, social network models have been applied to fraud detection. One study built networks linking claims, brokers, garages, policyholders and other participants, then applied specialized graph analytics to identify highly suspicious claims. These network-based approaches outperform traditional models, because they reveal hidden collusive structures that might evade conventional detection.

How Fraud Strategy Has to Change

Fraud Can't Be Evaluated One Claim at a Time Anymore: Most fraud controls were built to spot problems inside a single claim file. That approach works for opportunistic fraud, but it breaks down when activity is coordinated across people, vendors and claims over time. In network fraud, no individual claim looks especially suspicious. The risk only becomes clear when patterns emerge across claims, repair shops, providers or intermediaries. That reality is forcing insurers to connect signals across systems and teams rather than relying on isolated reviews.

Better Detection Still Requires Human Judgment: Advanced analytics and AI have made it possible to surface patterns that would never be visible through manual review alone. Network and relationship analysis can highlight repeated interactions and unusual clustering across participants, while machine learning can flag claims that deviate from expected norms. But technology does not replace investigation. These tools are most effective when they help experienced SIU teams focus their time on the cases that truly warrant deeper scrutiny.

Organized Fraud Rarely Stops at One Carrier: Fraud rings are adaptive and mobile. They move across insurers, jurisdictions and lines of business, learning quickly which controls are enforced and which are not. That makes fraud difficult to contain when each carrier operates in isolation. Industry collaboration, shared intelligence, and participation in fraud bureaus play an increasingly important role in disrupting organized activity. When carriers connect what they are seeing, repeat actors and emerging schemes become much easier to identify.

Prevention Starts Earlier Than Most Carriers Expect: As fraud becomes more organized, prevention matters as much as detection. Strong identity verification, consistent documentation requirements early in the claim, and training for frontline claims staff all reduce opportunities for coordinated fraud to gain traction. Clear communication around fraud consequences also acts as a deterrent. Over time, these measures lower investigative workload and help ensure that legitimate claims move through the system without unnecessary friction.

Network fraud is not a future concern for property and casualty insurers. It is already reshaping loss costs, investigative workloads, and customer trust. As fraud becomes more coordinated and less visible at the individual claim level, carriers that rely on traditional controls will continue to react after losses are locked in. Those that treat fraud as a networked risk can disrupt organized schemes earlier, protect margins more effectively, and reduce the burden on honest policyholders. The difference is not whether fraud exists. It is whether insurers can see it clearly enough and early enough to act.

A cyber incident hits a small business in a way that feels personal. The owner watches familiar tools fail, customers get locked out and schedules fall apart. They turn to their insurer looking for clarity, yet their expectations rarely match the process ahead. That disconnect shapes the entire claim and exposes where small businesses still misread cyber risk.

Many small operators still think that cybercriminals are carefully selecting their next target, but the reality is that automation has made it easier than ever for attackers to run broad scans across the Internet and strike wherever a shared account or outdated plugin gives them an opening.

It really doesn't take much. A rushed login, a forgotten update or a credential reused for convenience gives attackers the foothold they need. Those details rarely seem dangerous until the incident interrupts revenue, and the business owner realizes the breach reached everyday tools they rely on to run their business.

Misunderstandings around what triggers coverage

Confusion around coverage starts before a claim reaches the adjuster. Many small companies expect a cyber policy to function like a technical repair plan and anticipate quick fixes while assuming the carrier will take over the issue. Business owners also underestimate how fast costs build when dealing with investigations, data restoration and income loss that pop up in just the first hours of an incident.

The same pattern appears when incidents look minor from the outside. I've seen something as simple as a corrupted device wiping payment records for a food truck, and a damaged workstation forcing a photographer to cancel paid work. These events lacked the drama of national ransomware stories, but they still created lengthy downtime.

Tight staffing makes the fallout worse, because a single compromised login or failed device slows customer communication and sales. With no spare hands to absorb the disruption, small issues turn into long setbacks that strain the entire operation.

Where expectations diverge from what coverage provides

Cyber liability policies support recovery on two fronts.

• First-party coverage helps with investigation teams, data recovery, income loss and negotiation support during ransom talks.
• Third-party coverage addresses the fallout customers experience when their information becomes exposed.

Many small operators focus only on the technical failure and overlook how much involvement they will have once the claim starts. They may need to grant investigators access to certain devices or review which customer touchpoints were affected, and those responsibilities hit while they are already trying to steady the business.

On the flip side, executives hear from owners all the time who assume the carrier will fix the technical problem outright and feel blindsided when they learn how many steps sit between the first alert and a stable recovery. They do not anticipate the coordination needed to rebuild systems or manage customer notifications. That misunderstanding adds pressure to an already tense situation.

Security habits that stall during underwriting

Underwriting often uncovers a different kind of confusion. Many owners treat basic safeguards as add-ons rather than the foundation that keeps incidents contained, and controls like MFA or scheduled backups usually receive attention only when the application requires them, not when the business starts taking on digital risk. Once owners finally put those safeguards in place, they tend to keep them because the checks and alerts make daily operations steadier.

An office can cut risk by using a password manager and running brief phishing reminders with staff. A restaurant benefits when its reservation system restores from a recent backup instead of staying offline after a plugin failure. These steps rely on consistency more than technical skill, yet many owners delay them until an attack forces the lesson.

A clearer path forward

Cyber incidents reveal more about a business than the breach itself. They expose how prepared the team feels, how they handle uncertainty and how they respond when everyday tools fail. Insurance leaders watch this play out across industries, and those moments show the real gap between perception and reality for small companies.

Carriers cannot remove the stress that comes with a breach, but they can steady the path forward. Helping owners understand their role and their responsibilities changes how they navigate the experience. Cyber liability coverage gives them a path to continue operating during their hardest moments. Helping small companies understand that earlier reduces friction and sharpens the support they receive when an incident tests their systems.

When it comes to addressing connected extreme risks, companies have a number of choices, from building flexibility into their operations and talent, to combining quantitative analytics with qualitative 'storytelling' approaches to better identify and manage risks.

Understanding and quantifying risk tolerance is crucial to building financial resilience in a hyper-volatile world. However, many businesses fail to evaluate the critical pathways in which severe operational or revenue disruption could push them toward insolvency. A clear grasp of these scenarios and an organization's risk tolerance, combined with sufficient cash reserves, provides a buffer against unexpected financial shocks, insolvency or default.

By setting just enough cash aside, while considering the opportunity cost of doing so, a company will have the liquidity to continue operations and meet financial obligations, even in the face of hyper-volatility where significant and unforeseen climate or geopolitical risks are more likely to occur. With robust cash reserves an organization can, for example, quickly respond to a sudden increase in raw material costs without compromising financial stability.

Events like economic shocks, geopolitical disruptions and climate-related incidents can trigger one-off, unbudgeted losses. Financial analysis frameworks can be used to evaluate an organization's capacity to absorb such losses in excess of insurance in the context of its financial priorities, such as maintaining a budget, protecting credit ratings and preserving solvency.

Aligning reserve adequacy with defined risk tolerance thresholds provides a structured approach to evaluating an organization's financial resilience under stress by assessing the potential impact of those severe, low-probability loss events on financial performance.

We see more organizations managing hyper-volatility by building operational flexibility. Some practical examples an organization might consider include:  

Diversify revenue streams to spread across multiple products, services and markets to reduce dependence on any single source and mitigate the impact of sudden market changes or disruptions.

• Build supply chain redundancy to manage and mitigate the cascading effects of risks by having alternative options to continue operations in the face of disruption. With multiple suppliers and alternative production methods, a company can switch to a backup plan quicker than competitors if their primary supplier is compromised by climate, geopolitical or another driver of disruption.
• Digitalize to enhance operational flexibility. Cloud-based systems and remote work capabilities, for example, can help a company to continue operations despite the failure or destruction of physical infrastructure.

An organizational culture that values adaptability and innovation can prove vital for building financial resilience against hyper-volatility. When employees at all levels understand and prioritize the need to pivot and innovate, implementing and maintaining strategies that protect a company's financial health becomes easier. Continuous training, clear communication of an organization's risk management goals and known incentives for identifying risks proactively can help here.

Workshops on scenario planning put this in context and in practice: they leverage the problem-solving and critical thinking of employees beyond the risk functions, allowing them to better understand and then plan to respond to hyper-volatility.

By combining scenario testing, data and advanced analytics, an organization can create hypothetical scenarios that capture the cascading effects of multiple risks, such as natural catastrophe events or geopolitical tensions, and then analyze how these scenarios would affect their supply chain and financial performance.

Consider the 2018 and 2023 wildfires in Hawaii, which had starkly different outcomes. The 2018 fire was contained with minimal damage, while stronger winds and the presence of non-native grass species exacerbated the 2023 fire, leading to significant losses. By testing scenarios that consider multiple and nuanced environmental and climate-related factors, you can better identify and then address potential vulnerabilities.

Using advanced analytics to translate stress-testing findings into financial metrics can illustrate how a severe climate event, such as a drought, could materially affect a company's operations. Quantification of material impact helps make the business case appropriate resilience-building investments.

In a hyper-volatile environment, continuous learning and adaptation can prove essential. This is about reassessing an organization's risk management strategies regularly, using quantitative and qualitative methodologies and making adjustments as needed.

Staying informed about emerging risks and trends - combined with using advanced analytics to monitor real-time data on market conditions, such as fluctuations in commodity prices - means an organization can adjust financial modelling and cost structures more effectively in real time.

Acquiring a new insurance customer takes effort. Well-thought-out advertising campaigns, cold sales outreach, and personalized discounts are primary levers for building trust and expanding the customer base. This is a time-consuming and costly affair. That's why, over the last few years, customer acquisition costs have risen by 222%. And brands today lose an average of $29 per new customer, up from $19 a decade ago.

While the exact figures may vary, especially for a highly variable expense such as insurance, one thing is clear. All that effort can be undone by just one negative claim experience. A recent policy research report by Which? exploring consumers' experiences of the insurance claims process found that:

• 48% making a claim experienced at least one issue
• 28% consumers felt their insurer's actions were unjust
• 24% said they didn't understand why their claims got rejected

The message is clear: a single bad claim experience can erode trust, trigger churn, and damage a brand's overall reputation, especially in health insurance. A claim denial can occur due to errors in paperwork, missing documents, undisclosed pre-existing conditions, or other technical reasons. But there's no question that denial is brutal.

In this article, we explore why claims experience often matters more than customer acquisition, especially in health insurance, and how insurers can prioritize it.

The Reality of Claims Experience in Healthcare Insurance

Policyholders need clarity, trustworthiness, responsiveness, and timely claim settlements. Instead, they often get a claims stage that is marred by delays, manual paperwork, opaque communication, lack of explainability, and even denials. 

The TeleTech P&C Customer Satisfaction Survey highlights multiple factors, including how policyholders are treated, channel interactions, and the overall claims process. The most influential was, "my insurance company acted in my best interest."

Source: TeleTech Survey

This is the potential of a good customer journey, being there for the customer when they need you the most.

Where the Challenge Lies

It's not that insurers don't want to deliver superior customer experiences. However, that's just not possible today while operating on legacy CRM tools or spreadsheets. Processes are outdated, require repeated re-entry of customer data, are prone to errors, and are time-consuming. Also, these legacy, disparate systems offer no real-time insights into customer behavior or interactions, leaving insurers to guess based on historical patterns rather than available data, and customer behavior is dynamic and constantly evolving today.

The digital appetite is growing exponentially. Customers demand 24/7 brand availability. Customers want a self-service, AI-assisted portal. Instead of spending minutes on hold and then repeating every detail, they can now submit a claim through a conversational AI that's integrated with an insurance database and auto-populates the most relevant information. Also, this can help them check claim status in real-time, instead of calling up the agent to inquire about progress. This even empowers the employees, who can then focus on more innovative and complex tasks that machines can't replicate, like approving claims faster, offering empathy and transparency, and building stronger customer relationships.

AI-powered healthcare claims processing software can help. It streamlines claims processing, reduces administrative burden, and minimizes claim denials. With AI-driven insights and rule-based processing, insurers and providers can achieve a real-time, 360-degree integrated customer view, enabling them to take strategic initiatives to improve the customer experience.

Financial and Strategic Benefits of Prioritizing Claims Experience

Higher renewal rates and customer lifetime value (CLTV):

For health insurers, policy renewals constitute a significant source of recurring revenue. An insurer that is trusted in the market due to its higher claims settlement ratio, lowest rejection rate, and transparent communications will face far fewer challenges at renewal. This is far more effective than chasing individuals over calls, WhatsApp, or email, and it increases CLTV and brings predictable revenue streams.

Lower operational costs over time

Integrating new technology into a complex insurance process may seem complicated and costly, but a phased implementation can make it manageable, scalable, and beneficial. Unlike the traditional claims processing cycle, which involves manual paperwork, long wait times, and multiple checks, modern automated workflows lower operational overhead and reduce errors. AI-enabled claims management can reduce claims handling costs.

Less manual, repetitive work fosters employee satisfaction, and faster settlement leads to more satisfied customers. This makes claims operations both a service differentiator and a cost-saver.

Competitive differentiation and compliance

How do you build a competitive advantage in such a fierce market? Delivering outstanding claims experience can be a key. Most insurers are good at selling policies (acquisition), but they lack a strong retention strategy, especially in how to support a policyholder when they file a claim. An insurer that can offer a fast, fair, transparent, and smooth claims processing cycle would likely be a winner. Transparent claims handling can also help minimize customer complaints and fraud and optimize compliance and risk management.

Conclusion:

A healthcare emergency is already an emotional and physically exhausting experience for both the policyholder and their family. The last thing they want is a financial burden from an avoidable denied claim. The families shouldn't be chasing an insurance agent for a denied pre-authorization, especially when that policy covers the treatment costs. That said, incomplete/inaccurate patient information, healthcare plan changes, and submission errors can be among the other reasons for denied claims.

That's why a transparent, automated, and faster claims cycle can be the differentiator for businesses. Not only does it help boost operational efficiency by automating repetitive tasks, centralizing updated customer details, reducing data duplication, and increasing revenue streams, but most importantly, an AI-driven claims cycle helps an individual (the policyholder) in need. A smooth, caring, and empathetic claims handling means honoring the promise behind those sold policies when it counts.

The 2026 P&C insurance environment may be much easier to forecast compared with the last several years. Many experts have already said 2026 will be a year of profitable growth. Fitch suggests a combined ratio of 96% to 97%, and even the volatile homeowner market is anticipated to "stabilize." Similarly, it is obvious the new year will be filled with AI in insurance, building off widespread hype and numerous announcements of huge, multi-year investments by the likes of Travelers, Nationwide, GEICO, and Chubb. Just how much and deeply AI will affect insurance remains far less visible, with some areas of attention coming into focus, e.g., pre-binding, underwriting data, claims/FNOL analysis.

Looking back to "see" forward

We also know that each year can bring some unexpected and unwelcome surprises, such as the record-setting $40 billion-plus Palisades wildfires burning some 23,000 acres and everything in its path a year ago. On the flip side, not a single hurricane hit the U.S. Atlantic coastline in 2025. Global CAT losses still amounted to $107 billion in 2025, per SwissRe, but to put things in perspective, Hurricane Katrina in 2005 was around $105 billion alone. Severe CAT exposure has become more visible and thus accepted. In turn, risk models have presumably adjusted over the recent years, and, optimistically, the industry is better prepared.

Throughout 2025, M&A remained vibrant, with insurtech funding just over $1 billion per quarter and an increase in early-stage funding toward the end of year, per Gallagher. A closer look at some specific trends:

• Commercial rate softening, with much variation, e.g., property lines down, commercial auto up and workers' compensation flat
• New car sales slumping by 7.5% at year-end
• Car loan terms elongating beyond 60 months and car loan payments reaching a record of $760 monthly average, per JD Power
• Total loss auto rates rising, to 23%
• Overall auto claim repair volumes declining roughly 8%
• Deductibles for both auto and home climbing, shifting the financial burden from insurer to consumer. According to a study by MATIC, home deductibles are up 22%

As we look to 2026, the following trends help bring rationale for how we see things shaping P&C insurance into a year of progress and collaboration, including:

• Accelerating digitization
• Climate change
• Pressure on globalization
• Rising economic and social inequities
• Major demographic shifts
• Layoffs

AI in Insurance

Investments and practical insurance industry applications of AI will continue to expand even as a large number of AI startups fail and regulators try unsuccessfully to catch up to developments.

• AI will eliminate even higher numbers of less skilled employee positions, including transactional, customer service and document management. At the end of 2025, Chubb announced "radical" cuts of 20% over the next three years due to AI deployment. It is highly likely that other carriers will follow suit.
• A new breed of AI entrepreneurs will emerge to invent highly specialized consumer services delivering instant hyper-personalized gratification for information, retail therapy, mental wellness and unique "experiences."
• AI-enabled photo inspection will gain greater adoption across the insurance, automotive and transportation segments using computer vision and machine learning to automatically analyze images for defects, anomalies, damage, or authenticity, replacing or assisting manual visual operations. This technology will be applied across various industries including insurance to enhance efficiency, consistency, and fraud prevention.

Other Key Factors

Affordability will reverberate beyond a broad consumer/political cost-of-living issue, circling back to P&C insurance where unaffordability chants arguably began. Cost of coverage, availability, pricing and rating methodology will draw even greater attention from consumer groups and regulators. Protection gaps and total cost of home or vehicle ownership will become primary concerns replacing 2024/25 inflation and supply chain worries.

Sustainability will gain adoption across the insurance value chain, especially the North American ecosystem, influenced by global re/insurers. Areas of early focus will include risk and claims management such as auto physical damage and property. Sustainable insurance will reduce risk, develop innovative solutions, improve business performance, and contribute to environmental, social and economic sustainability. Lessons learned from consecutive catastrophic events may serve as a tipping point, taking holistic approaches to predict, harden and prevent.

Climate risk modeling will gain energy. Demand is high and growing for accurate, usable climate information, particularly data that can help assess risk more accurately and contextually. This will drive carriers and others to probe every level of risk, from neighborhoods exposed to more frequent flooding, and to test if proposed atmospheric cooling approaches can work safely, if at all. Interest and investments in climatetech that can benefit insurance will grow significantly.

Cyber threats and fraud losses will continue to expand as digitalization spreads around the world. Recent cyber claims frequency trends remained low while severity increased, presenting the insurance industry with a huge—yet hugely challenging—opportunity.

Insurtech consolidation will accelerate as partnerships and acquisitions become de rigueur and single-point, stand-alone solutions continue to lose favor. The future of insurtech is shifting from rapid disruption to sustainable, AI-driven integration, with the market projected to reach up to $254 billion by 2030. Key trends include AI-powered automation for claims and underwriting, hyper-personalization, embedded insurance, and a focus on profitability over pure growth. Agentic AI platforms will automate routine tasks, potentially cutting outsourced insurance roles in half by 2028. AI will also enhance risk assessment and, in some cases, replace traditional underwriting.

Embedded insurance will continue to emerge as a significant distribution channel as discreet insurance offerings are packaged with the related product/service purchase at the point-of-sale in a singular transaction. Auto insurance packaged with new and used cars will grow as OEMs and dealers seek new profitable revenues.

Open platforms and marketplaces will continue to proliferate, and closed systems will face greater headwinds. Core insurance system platforms (e.g., Guidewire, Duck Creek, Majesco) now host hundreds of popular third-party product and service providers supporting claims, policy administration, billing and payments. Even the leading auto claims and repair information providers (e.g., CCC Intelligent Solutions, Enlyte/Mitchell and Solera/Audatex) have begun to pivot from proprietary closed systems to partnerships with emerging physical damage solution providers.

Consulting firms will restructure for agility, such as moving toward "one firm" models to blend technology and consulting, exemplified by PwC's 2024 internal leadership changes. Major firms like McKinsey, Accenture, and the Big Four have implemented layoffs amid shifting demand for services. AI is cited as a major driver, with a growing percentage of McKinsey's work involving AI-related projects and ultimately affecting the need for traditional roles.

Tech talent will be at a premium. All companies will scramble to retrain, upskill and upgrade staff to fully leverage new and emerging technologies. Change management principles will be dusted off and applied to the numerous influences AI will bring to enterprises seeking to excel. A recognition of the importance of people leveraging AI as much as replacing work functions will continue.

Regulatory risk management will require agility in a fragmented landscape. Navigating a changing regulatory patchwork will require investments in legal expertise and compliance infrastructure, which can drive up operating costs. Some insurers will likely cut their losses and focus on less risky markets. That could increase return on investment for organizations that try to take a broader approach with a longer strategic horizon.

Caveats/The Future

If we have learned anything from recent history it is that the future is inherently uncertain. Not all of our predictions will materialize. Black swan events may occur, reshaping markets, nations and the insurance industry in dramatic, unpredictable ways. But we are confident that most of what we have forecast will become reality. Either way, our projection of optimism for a healthy and vibrant P&C insurance industry in 2026 tops the list.

We look forward to seeing the industry move forward, making progress through technology and collaboration. 2026 will be another exciting year for the industry in both expected and unexpected ways.

Moments after the Baltimore Ravens kicker missed a near-gimme field goal attempt with no time left and let my Pittsburgh Steelers escape with a two-point victory that won them the AFC North and got them into the NFL playoffs, ESPN sent out a notification. The headline read, "The North Belongs to Ravens."

Oops. 

I'm sorry I was too busy whooping and hollering at my TV to quickly click on the notification, and ESPN had fixed the story by the time I got to it, but I can imagine the tone it took about my Steelers. After all, I've read all the vitriol aimed at the Ravens. 

Ravens head coach John Harbaugh would have been a hero in Baltimore if he'd won the game against his archrivals and snuck into the playoffs, but now there is even speculation he will lose his job. He had his team in a position where it was so obviously going to win that I had my finger on the remote, ready to turn the TV off the moment the ball went through the uprights, so I could go out into the yard and scream. But his kicker missed the field goal, so, boom, let's get rid of the bum.  

The coverage is not only wrong-headed — as much as I'm happy to see people beat up on the Ravens and Harbaugh — but demonstrates the sort of mistake that makes executives, including in insurance, evaluate innovation efforts poorly.

The problem is that we don't think in percentages, or at least not well. We may know that the Baltimore kicker had a nearly 90% chance to make his 44-yard field goal try, but we don't quite get that the percentage means he'll miss one time in 10. We know it, but we don't really believe it. So when the kicker misses, we just see failure and look for someone to blame. 

The Yahoo Sports writer acted as though the Ravens failure to make the playoffs was foreordained, even though they began the season as one of the top-ranked teams in the league. His article began: 

"From the beginning of the Baltimore Ravens’ season, when they had an epic collapse and lost to the Buffalo Bills, right to the end when their season ended with a loss to the Pittsburgh Steelers, nothing was good enough." He added that the Ravens "will be searching for answers after a season that went horribly wrong. There will be immediate questions about John Harbaugh’s future as the team’s coach."

The Steelers and head coach Mike Tomlin would have come in for exactly this sort of treatment if the field goal had been good, even though the Steelers had even a stronger gripe about the kicking gods being against them. The Steelers were only two points ahead, and thus in danger of losing to a field goal, because their kicker had missed an extra point with 55 seconds to go in the game, after not missing one all season. The pressure on Tomlin isn't entirely gone, given that he hasn't won a playoff game in eight years, but there will be a lot less of it in the off-season, especially if we beat the Texans in Pittsburgh on Monday night. 

The Steelers even came in for misconceived criticism despite winning the game. I always liked Trey Wingo when he was an analyst on ESPN, but he posted a Tweet that was downright silly. He said Tomlin made a terrible decision by going for a field goal at the end of the first half rather than trying for a touchdown from the Ravens one-yard line. He tried to back up his claim, but he was clearly just Monday morning quarterbacking — because the Steelers failed to score a touchdown, they shouldn't have tried.

The truly goofy part of his argument was his statement that, had the Steelers converted a chip-shot field goal, they would have been five points ahead of the Ravens at the end of the game, not two, and wouldn't have had to worry about a field goal. But come on. If the Steelers had scored a field goal, the dynamics of the game would have changed. The coaches would have made different decisions, and the players would have done different things in the different circumstance. You can't just take the three points and add them to their score for the rest of the night. 

The other part was nearly as bad. He offered an adage: "You always take the points." But the NFL has moved beyond a lot of adages and started to apply real, live math to questions such as when to go for it on fourth down, as I wrote a year ago. And here's the math:

A field goal would have been almost a 100% chance, so the Steelers could have counted on almost three points. The attempt at a touchdown had a slightly better than 70% chance, based on league averages, and an extra point is almost a lock. 70%-plus of seven points is roughly five points. Yes, the Steelers' attempt was embarrassingly incompetent, but I'm still going to take a likely outcome of five points over one of three points in almost every circumstance. 

This is the Ravens we're talking about. I need every edge I can get.

What does this mean for insurers?

Insurers need to think more like venture capitalists and less like Yahoo Sports or Trey Wingo. Venture capitalists not only know that 90% of the startups they invest in will fail but act on that knowledge. If they think an area is promising, they make a number of bets rather than assuming they're going to pick the one winner. They don't label their entrepreneurs as losers just because they lose — if you had a 90% chance to win but lost, they count that as a win and blame the circumstance, not you. As a result, VCs often invest in serial entrepreneurs, who have either previously failed or only had modest successes. 

A line from the mother of a girl who ski raced with my younger daughter has stuck with me. Stephanie, who build a financial software business she sold for $2.5 billion, told me: "I like people with scars."

Insurers also need to think of early innovation efforts as opportunities to learn, rather than trying to immediately shape them into pilot projects designed to scale and go to market. As I've been saying for almost 30 years now, the key is to Think Big, Start Small and Learn Fast — the latter two points meaning that you need to do lots of inexpensive projects, even though they're in the service of a grand vision, and move on quickly to the next set of tests once you've learned what you can. The vast majority of these projects won't get anywhere near the market, but they aren't failures if you've learned something important.

I'm hardly arguing for no accountability. There are still bad ideas, and they may be executed poorly by people who should no longer be part of your organization. I'm just arguing that we all take a more sophisticated view of success and don't decide that failing to convert a 90% chance merits firing — even though I'd love to see Harbaugh out of Baltimore.

Go, Steelers! Beat those Texans!

Cheers,

Paul

Paul Carroll

You’ve written for us about the need for hyper-personalization in life insurance. Would you start us off by describing what that looks like in practice, as well as how it differs from how life insurance has historically been handled?

Brooke Vemuri

I've been in the life insurance business for 23 years, working in both technology and operations, so I've seen firsthand the evolution from moving physical forms around the building—we had folders, and we put them in carts, and we drove them around to our underwriters—to where we are today. We've crossed a lot of hurdles over the last 20 years.

Now we're in a place where we have intelligent automation and a digital application process. Over the past few years, this has meant having an event-driven, rules-based system that allows us to capture the right application questions, then run rule sets underneath, call third-party data, and do all the things we need to amalgamate and come together on a decision. That change was hard to get across the line, but we've arrived and are doing very well as a result.

The next evolution is moving away from considering the application as one static process or one way to get to an outcome. Instead, we're moving toward tailoring the experience by distribution, by agent, by customer. Personalization means accommodating the way our agents and agencies like to do business.

Just to give you a small example: We have some agents who say, "Don't waste my time collecting beneficiary information. I'm going to get you everything I need to get a decision on the case. Then once I have a decision and the customer accepts, I'll gather that beneficiary information—it's more of an administrative piece of work because it's not influencing the decision." I have another distribution partner that says, "We start with beneficiaries. We connect the sale to the reason why you're making the purchase."

That's a high-level example on the easier side of personalization—how do we tailor or reorder the journey? 

Next comes determining how many questions we ask based on the product, and how we tailor that experience based on the product, the partner, and the customer. You start to get all these different variations of how you would flow a digital application process to collect the right information at the right time, make the right decision, and end up with a case you can place in force—in a way that works with every agent and partner you have.

You can start to see that there are going to be lots of permutations and combinations of how all of that evolves and comes together. From a technology perspective, that's all about creating the right, flexible architecture to make that happen, to allow that configuration, and to support our agents in the way they want to sell the business.

Paul Carroll 

Is this personalization extending beyond the sales process to policy offerings and features, as well?

Brooke Vemuri

Yes, that's a good question. Over the last 20 years or so, you typically entered the journey when you already knew what product you wanted. For example, "I know I want a term 20 policy. It's for a 50-year-old male with two children.” Now we're heading toward a different approach—someone starts the journey as that same person, but they really don't know what product or combination of features they want until they move through the journey.

Whether that means recommendations based on how many beneficiaries you have, what stage of life you're in, what your income levels are, or what job you have—as the system starts to understand who the customer is in the journey, we can start to make the right recommendations for a product.

I think you're going to see both tailoring of products and features. One of the things we're working on is how to come out with an offer for every applicant. Because in a lot of cases, there are still declines in our environment, even on term business. So how do you enter the process saying, "I have a desire to have life insurance" and be sure to end up with something—whether that's an accidental death product or a final expense product?

So yes, you're going to start to see tailoring of offers, cross-sell, counter-offers—that's what we're calling it. How do we come up with another product that might be viable for both the customer and the life insurance company so that, at the end of the journey, you still get some product that is available to you.

Paul Carroll

How is AI being incorporated into this process, particularly in terms of gathering customer information and providing real-time recommendations to agents during customer interactions?

Brooke Vemuri

Intelligence is coming into play primarily in our agent-facing capabilities. Our journey encompasses both customer-agent interactions and internal, employee-facing systems. As people interface with the system, it makes certain recommendations to the agent based on how the journey is progressing. 

You might not do that directly with a customer, though. If a customer is going through the process unassisted, it's a little bit more complex for them to navigate some of those things. So our thought process is a lot like TurboTax—in the upper right-hand corner as you progress, it shows what you're accruing, what we know now, and where we might go in this journey, and starts to forecast and give options for next best steps.

For now, we’ll just focus on agent-facing capabilities, because that feels more like advisory-type activities. That's at least our early thinking.

Paul Carroll

What innovation can we expect to see over the next few years?

Brooke Vemuri

I think we'll be very much focused on what we're talking about now, which is tailoring our apply processes—for both customer and agent. We're looking at getting some cues or indicators into the journey about how a case is tracking and how agents might handle what's going to happen in terms of that sale, whether it's going to be a decline or whether we can pivot to another product.

I think you're going to see all of that in the next two to three years—tailoring the counter offers or the alternate offers, tailoring the journey to align with how that agent wants to sell their business.

Beyond that, we're imagining a lot more variation on the product. We're going to see more things in terms of what we call table ratings today: How do we get more price variability in the product so we can accommodate more and more applicants? I think that's probably on the three- to five-year horizon.

Paul Carroll

What trends are you seeing around younger generations being more interested in living benefits rather than death benefits?

Brooke Vemuri

That's a good point. Things are being added on and tailored into the product that allow more utility out of the product. It's not a policy that you print and put on the shelf and wait till you die, and then someone gets the benefits of it. It's about how the policy can give you benefits while you're living—tapping into that face amount for critical illness or accidents and things like that.

I agree that living benefits are giving more value to the term product. I mean, there's still a need for your basic term product out there that has no bells and whistles. But to reach some of the younger generations, we're finding that living benefits have been valuable for them.

Paul Carroll

Life insurance policies represent long-term commitments—20 years for term policies, often longer for permanent ones—making it challenging to validate underwriting assumptions quickly. How is the recent industry shift toward fluidless evaluation and fewer application questions working in practice?

Brooke Vemuri

It does take years to know if your assumptions are right. I think the biggest value of the data right now is being able to look at historical data and model it in a way that makes us more comfortable with innovations—whether that's leveraging other third-party evidence sources or forgoing exams altogether, like with fluidless underwriting.

We're using a combination of historical data—looking at how it would perform based on our back book of hundreds of thousands of cases at this point—and other third-party evidence that we can leverage, like medical claims data and labs data. We're leaning heavily into claims data specifically.

It's really about putting those pieces of the puzzle together alongside the self-declarations from the application so that you have a complete picture to underwrite from and make a decision. We're going after alternative evidence sources to eliminate some exams while also looking retrospectively at the data to see what it's telling us. Through our modeling, we can apply new assumptions to that data, see what it might look like, and price accordingly.

Paul Carroll

What other trends should our readers be aware of in life insurance?

Brooke Vemuri

I think you've probably captured a lot of them. There's going to be a lot of tailoring—tailoring on the journey and tailoring on the products, tailoring on the prices, tailoring on the variation of features and capabilities that exist in the term product space.

How do we alternate out of term if we need to? How do we counter out of that if we need to? I think that's the real progress on the horizon.

Paul Carroll

Thanks, Brooke.
 

About Brooke Vemuri

Brooke Vemuri is vice president, IT and innovation at Banner Life and William Penn. She leads people and cross-functional teams to reimagine the future of life insurance from lead generation, through apply and underwriting, to offer, pay, and in-force. Her team drives transformation and change to the business and distribution through the development and execution of business propositions focused on growth and cost reduction through a digital business strategy.