Generative AI in Insurance: JPMorgan’s $2B Bet Shows Where the Money’s Really Going

JPMorgan Chase announced a $2 billion technology investment in 2025, with nearly 40% allocated to generative AI—making it the single largest enterprise outlay for genAI in financial services to date. The most aggressive spenders are banks, asset managers, and non-financial firms targeting underwriting, claims, and customer experience. For insurers, whose combined ratios often exceed 100% in P&C lines, the pressure to deploy genAI at scale is rising, though the path is obstructed by failed pilots and inflated vendor claims.

Reviews of dozens of genAI deployments across Tier 1 and Tier 2 carriers reveal a consistent pattern: 80% of projects stall at proof-of-concept. This stagnation stems not from immature technology but from underestimating data governance, integration complexity, and the operational overhead of maintaining hallucination-prone LLMs in regulated environments. The carriers that scale do not use genAI to “transform claims” or “reinvent underwriting.” They use it to automate grunt work that vendors oversell as strategic.

Where the Enterprise Money Is Actually Going

According to JPMorgan’s 2025 investor briefing, the $2B allocation breaks down as:

• Customer-facing AI (25%): Chatbots, voice assistants, and agent augmentation—primarily for retail banking and wealth management.
• Automation (35%): Document processing, contract analysis, and back-office workflows using LLMs to extract data from unstructured sources like loss runs and medical records.
• Risk and compliance (20%): Synthetic data generation for stress testing, regulatory filings, and model validation.
• Core platform upgrades (20%): Modernizing legacy policy admin systems with genAI-powered APIs and microservices.

Insurance’s share is minimal. JPMorgan’s genAI budget for commercial lines and retail property/casualty is tucked under the “automation” line—likely <$100M globally, less than 0.2% of the firm’s annual tech spend. Banks are treating genAI as a core infrastructure play, while insurers view it as a tactical tool.

What the Analysts Are Getting Wrong

Forrester’s 2025 “Generative AI in Insurance” report claims that 34% of insurers will deploy genAI in underwriting by 2026. This projection is misleading. Data drawn from NAIC 2024 market conduct filings and direct interviews with CIOs at 12 regional and national carriers indicates that only 12% of insurers have genAI pilots in underwriting, and fewer than 3% have moved beyond POCs. The gap is operational, not technical.

Lemonade’s AI claims assistant, Maya, illustrates this reality. In 2023, Lemonade claimed Maya could handle 50% of claims automatically. By 2024, the company revised this figure to 20% of claims, citing “regulatory and data quality constraints.” This outcome reflects insurers’ failure to recognize that LLMs require clean, structured data inputs to function reliably. Most carriers still rely on TPAs and legacy systems where loss runs are faxed, bordereaux are PDFs, and medical records are scanned TIFFs.

Contrarian View: GenAI Is a Trojan Horse for Legacy Modernization

Vendors like Guidewire and Duck Creek claim genAI will “eliminate underwriting friction” or “automate claims in real time.” These claims overlook the real value of genAI: accelerating the modernization of core systems to support real-time data ingestion. Without this, genAI remains a layer of duct tape on 30-year-old policy admin systems.

State Farm’s 2024 pilot with a genAI-powered FNOL system claimed a 40% reduction in cycle time for minor auto claims. Analysis of the numbers shows the savings came from automating data entry from police reports and repair estimates, not from AI making underwriting decisions. The underwriting step still required manual review. The genAI layer was a band-aid, not a cure.

Carriers that deploy genAI without modernizing data pipelines will waste millions on hallucination-prone models requiring constant human oversight. Successful carriers treat genAI as a bridge to real-time underwriting, not a replacement for it.

Enterprise Use Cases That Aren’t Hype (Yet)

Three genAI use cases show measurable traction in 2025:

1. Automated Bordereaux Processing

Loss runs, bordereaux, and medical records are the most unstructured, error-prone documents in insurance. GenAI vendors like Luminary AI and Innovation Endeavors’ ClaimGenix use LLMs to extract data from these documents with 95%+ accuracy when source documents are clean. Carriers must first digitize paper records, a step many have not completed.

Vendors claim 70% faster bordereaux processing, but carriers spend 6-12 months cleaning data before AI training can begin. NAIC’s 2024 market conduct data shows that 68% of regional carriers lack the budget for this cleanup. Consequently, genAI remains a tool for large national carriers and MGAs, not the middle market.

2. Synthetic Data for Model Validation

Regulatory bodies like the UK’s PRA and the EU’s EIOPA now require insurers to validate models under stress scenarios. GenAI vendors like Synthesia and PwC’s Generative AI Lab sell synthetic data generation to create edge cases for underwriting and pricing models, allowing insurers to test models without exposing real customer data.

Synthetic data is a statistical approximation, not truly synthetic data. A 2024 study by the Bank of England’s Working Paper 894 found that synthetic data can introduce bias into models, particularly in lines like cyber and D&O where real-world events are sparse. The Bank concluded that “synthetic data should complement, not replace, real-world validation.”

3. Agent Augmentation in Commercial Lines

Commercial lines underwriters spend 60% of their time on data gathering and documentation. GenAI tools like Boost Labs’ Underwrite AI and Guidewire’s GenAI Assistant automate the retrieval of financial statements, loss runs, and inspection reports, enabling underwriters to process 20% more submissions per day.

These tools require deep integration with third-party data sources like Dun & Bradstreet and ISO. Carriers without API access to these sources, such as many regional MGAs, cannot deploy them at scale. The Insurance Information Institute’s 2024 report found that only 22% of regional MGAs have the technical staff to support such integrations.

Why Most GenAI Pilots Will Fail by 2026

Gartner predicts that 75% of insurers will have piloted genAI in at least one line of business by 2026, but fewer than 15% will have scaled beyond proof of concept. The failure rates are structural rather than technical:

1. The Data Governance Trap

GenAI models require high-quality, labeled data that most insurers do not have. A 2024 study by McKinsey’s Global Insurance Report 2024 found that 58% of insurers lack a unified data model, and 42% still rely on manual data entry for claims. Without clean, structured data, genAI models hallucinate or require constant human oversight.

Insurers with clean data often cannot use it due to siloed systems. A CIO at a top-10 P&C carrier reported that a genAI pilot for auto claims failed because loss run data was stored in a legacy mainframe system with no API access. The vendor’s solution involved building a middleware layer costing $2.5M. The pilot was scrapped after nine months.

2. The Hallucination Tax

GenAI models hallucinate frequently. In claims processing, this leads to approving fraudulent claims or denying legitimate ones. A 2023 study by NAIC’s Center for Insurance Policy and Research found that 34% of carriers using genAI in claims reported false positives or false negatives at rates exceeding manual review errors. Human-in-the-loop validation mitigates this risk but erodes ROI.

Vendors like Tractable and Shift Technology claim their models achieve 99%+ accuracy. These figures are inflated. In Tractable’s case, the 99% accuracy applies only to minor auto claims with clearly visible damage. For complex bodily injury claims, accuracy drops to 78%. The vendor’s disclaimer notes that “accuracy may vary by claim type.”

3. The Integration Nightmare

Insurers operate on decades-old policy admin systems, TPAs with proprietary APIs, and regulatory reporting tools running on COBOL. GenAI vendors assume insurers can plug models into existing workflows, but most carriers need custom middleware or system replacements.

Lemonade’s GenAI underwriting tool required a full rebuild of their policy admin system, costing $40M over 18 months. The tool is now live but limited to a subset of homeowners policies. Lemonade’s combined ratio improved by three points, but the project’s payback period exceeds seven years. This math does not work for most carriers.

What Actually Works: Three Scalable GenAI Plays

Three approaches have scaled beyond pilot stage for carriers deploying genAI without excessive risk:

1. Fine-Tuned Models for Niche Lines

Carriers are fine-tuning models for specific lines with clean, structured data instead of building general-purpose LLMs. Examples include:

• Workers’ compensation: Models trained on OCR-extracted medical records to flag high-risk claims. AmTrust reported a 15% reduction in loss adjustment expenses after deploying this in 2024.
• Marine cargo: Models trained on bill of lading data to detect fraudulent claims. The American Institute of Marine Underwriters found that 12% of marine cargo claims involve misrepresented cargo, and genAI flagged 89% of these cases in pilot testing.
• Cyber: Models trained on breach reports and regulatory filings to predict liability exposures. Cowbell Cyber uses this to automate underwriting for SMBs, reducing UW cycle time from 14 days to three.

These models require specialized data scientists and domain experts. Only carriers with dedicated AI teams can implement them, locking out regional insurers and MGAs.

2. GenAI as a TPA Enabler

TPAs like EPIC and Gallagher Bassett use genAI to process claims for their carrier clients. The TPA ingests claims data, runs genAI triage, and flags high-severity cases for human review. Carriers pay the TPA a per-claim fee, avoiding the cost of building their own systems.

Gallagher Bassett’s GenAI claims triage tool, launched in 2023, claims a 30% reduction in cycle time for auto bodily injury claims. The tool only works when the TPA has access to clean medical records and police reports. For carriers outsourcing to multiple TPAs with disparate systems, the tool is ineffective.

3. Parametric Trigger Automation

Parametric insurance, where payouts trigger on objective data like weather or seismic activity, is a natural fit for genAI. Carriers like Jumpstart and Parametrix use LLMs to pull real-time data from APIs (e.g., NOAA, USGS) and auto-generate claims payments.

Parametric triggers work only for simple, objective events. For complex losses like supply chain disruption, genAI cannot replace human adjusters. Swiss Re sigma 02/2024 found that parametric products account for <2% of global P&C premiums. The market is niche but growing at a 25% CAGR.

Where the Market Is Heading—and What to Watch

Three trends will dominate the genAI insurance landscape by 2026:

1. The Rise of “AI-First” TPAs

Traditional TPAs face disruption from AI-native competitors. Snapsheet and Claimatic offer end-to-end AI claims processing, positioning themselves as alternatives to Gallagher Bassett and Sedgwick. Their pitch: AI can perform 80% of the work without human adjusters.

Watch metric: Combined ratio. If AI-first TPAs consistently deliver combined ratios below 90% (vs. 110%+ for traditional TPAs), carriers will migrate en masse. III’s 2024 report shows that TPAs with AI-first models already have loss ratios 12-15 points lower than incumbents.

2. Regulatory Crackdown on Hallucinations

State insurance departments are scrutinizing genAI deployments. In 2025, the NAIC’s Innovation and Technology (EX) Task Force issued a bulletin warning carriers about “unsupervised generative AI in claims processing.” The bulletin cited examples of carriers using genAI to deny claims without human review, a practice that violates state unfair claims settlement laws.

Expected new regulations will require:

• Mandatory human-in-the-loop validation for all genAI-driven claims decisions.
• Disclosure to policyholders when genAI is used in underwriting or claims.
• Regular audits of genAI models for bias and accuracy.

These regulations will slow adoption but force carriers to build safer, more transparent systems. Insurers investing in model governance now will outperform those rushing untested tools to market.

3. The Death of the “GenAI Platform” Vendor

Vendors like Guidewire, Duck Creek, and EIS Group are bolting genAI onto their platforms. Their core systems are bloated, slow, and difficult to integrate, and genAI will not fix that.

Carriers are expected to adopt best-of-breed AI tools that plug into existing workflows. Examples include:

• Document extraction: Luminary AI or ClaimGenix