AI Patents in Insurance: A Concentrated Race With P&C Out in Front
Artificial intelligence patenting in insurance has become a high-stakes, highly concentrated race. New data from Evident's Insurance AI Patent Tracker shows activity is still 30% below the 2020 peak, yet the winners are clear: a handful of U.S. property and casualty carriers are outpacing everyone.
Since January 2023, 30 major insurers across North America and Europe filed 166 AI-related patents. The bulk comes from a small group of U.S. firms. State Farm, USAA and Allstate account for 77% of insurer AI patents filed over the past decade, and P&C carriers are responsible for 89% of filings overall.
"Patents offer a rare window into where insurers are placing their biggest bets on AI," said Alexandra Mousavizadeh, Co-founder and CEO of Evident. "This data shows that innovation is overwhelmingly being driven by a handful of US firms, especially in P&C. While the volume of filings remains modest compared to banking, we're seeing a sharp uptick in generative and agentic capabilities. The IP focus is moving from protecting past systems to enabling future ones."
Why P&C Leads
P&C carriers have a built-in advantage: telematics, IoT-driven risk monitoring and sensor-heavy systems tend to meet patentability standards more cleanly in the U.S. and Europe. When an invention shows a clear technical effect-especially tied to hardware, data capture, or real-world control-claims are easier to prosecute.
If you operate in this space, you already know the playbook: data ingestion at scale, model-driven decisioning, and automated actions tied to a vehicle, property, or device. Those ingredients translate into patentable subject matter more often than pure process automation. For reference, see the EPO's guidance on AI and ML patentability standards here.
GenAI and Agentic Systems Are Climbing
Overall filing volume hasn't bounced back-Evident notes aggregate activity sits roughly 30% below the 2020 high. But inside those filings, the mix is shifting hard toward generative AI. GenAI patents moved from 4% of insurer AI filings to 31% in 2023, mostly around customer service and claims automation.
Agentic AI-systems that can take autonomous actions for users-is just starting to show up, with only three insurers filing so far, led by USAA. That early movement matters. As soon as these systems touch adjudication, payments, or policy changes, expect a wave of defensive IP and new prior art to navigate.
Where Insurers Are Applying AI Today
The filings give a clean look at how leaders are embedding AI across the value chain:
- USAA: generative AI to clarify aerial imagery for property damage assessment.
- State Farm: machine learning for claims triage and autonomous vehicle fault analysis.
- Allstate: in-vehicle AI assistant to automate parts of claims and support behavior-based discounts.
- Swiss Re: predictive analytics for medical data and anomaly detection.
- MassMutual: interpretable underwriting and AI-driven document tagging.
- Liberty Mutual: generative AI for engineering release notes.
- Zurich Insurance Group: systems that convert user-entered addresses into clean, structured datasets.
Claims and underwriting remain the center of gravity. The sector has filed over 300 AI patents in these areas-roughly double the next tier of focus: customer service, risk modeling and pricing.
Strategic Crossroads for Carriers
Evident frames the moment as a choice. Either a few frontrunners, especially in U.S. P&C, use IP to lock in data and automation advantages-or patent activity becomes a broader signal that more carriers are moving from pilots to production-scale AI.
"The insurance sector is at a crossroads," added Mousavizadeh. "Either patents remain the domain of a few frontrunners, or they become merely a signal of broader competitive intent. As generative and agentic AI rework the value chain, insurers will need to decide whether to build IP defensively or lead from the front."
What To Do Next: Practical Moves for Insurance Leaders
- Audit your IP vs. roadmap: map active and planned AI systems (claims, underwriting, FNOL, SIU, customer ops) to potential patent families. Decide what to patent vs. keep as trade secret.
- Prioritize "technical effect": filings tied to telematics, computer vision, sensor fusion, or automated controls tend to fare better under subject matter rules.
- Define an agentic stance: identify low-risk, high-impact actions agents can take (e.g., pre-fill, triage, routing) and file early where differentiation is real.
- Hedge with data: your proprietary datasets and labeling pipelines can be as defensible as models. Document data provenance and quality controls in filings.
- Stand up governance: align claims, legal, security and model-risk teams. Follow inventorship and AI-assisted drafting rules set by regulators like the USPTO guidance.
- Track competitors weekly: monitor grants and applications to avoid dead-ends and spot licensing or cross-lab opportunities early.
Bottom Line
AI patenting in insurance isn't broad; it's concentrated, deliberate and moving toward generative and agentic capabilities that tie directly to claims and underwriting. If you're serious about AI at scale, treat IP as a product decision-file where it creates leverage, skip where it doesn't, and keep your data advantage front and center.
If your teams need to skill up on generative, agentic and automation skills relevant to insurance operations, explore curated training by role here.
Your membership also unlocks:
