AI fraud detection and claims processing reshape $7 trillion insurance market
In 2024, insurance carrier Allianz received a damage claim with a photo that looked credible. The fraud team investigated the customer's social media and found the same image-edited to show fake damage. The claim was denied.
This scenario is becoming routine. Fraudsters now use AI tools to manipulate images. Insurance companies are fighting back with their own AI systems to verify damage and stop payouts on false claims.
Ravin, an Austin-based startup, built software that lets customers video-scan vehicle damage with their phones. A vision model assesses severity and flags manipulated images by collecting metadata about location and timestamp. The approach makes it harder to submit edited photos as evidence.
"Insurance companies increasingly accept images as evidence," said Eliron Ekstein, Ravin's founder and CEO. "With our technology, you can't just upload a set of images. You need to perform a scan that will take the images for you."
Automation addresses labor-intensive workflows
The insurance industry has relied on manual processes for decades. Underwriters assess risk, draft contracts, and process claims-all labor-intensive work that slows payouts and frustrates customers.
Corgi Insurance, a San Francisco-based carrier, became the first AI-native insurance company to win U.S. regulatory approval in July 2025. It scaled to $40 million in annual recurring revenue within months by automating text-heavy workflows.
"Most of our competitors employ north of 40,000 people that do very repetitive workflows," said Nico Laqua, Corgi's co-founder and CEO. "In our case, we automate as much of that as possible."
Corgi uses generative AI and LLMs to interpret contracts, generate regulatory reports, and assess claims. These tasks are language-based-a natural fit for large language models.
The speed advantage matters. When a house burns down at midnight, customers need settlement immediately. Humans cannot respond that fast. Corgi's system can.
Compute costs rise with regulatory demands
Running AI in insurance isn't cheap. Regulators demand low tolerance for bias and hallucination in a financial product. Corgi addresses this by running multiple models in parallel-one model checking another's work.
"Supervisory models have gotten quite good so you can use models to oversee other models with anything that is super, super sensitive," Laqua said.
This approach multiplies compute costs by a factor of four per inference call. Add the volume of text processing-contracts, forms, regulatory reports-and the bill grows quickly.
"We use a lot of tokens, and then we need to double and triple and quadruple-check all of the work that we do because we're selling a financial product, and it needs to be correct," Laqua said. "So that's just expensive."
Ravin processes 2,000 videos daily. The sensitive data those scans capture requires regionally compliant cloud infrastructure to meet data residency rules.
Market growth accelerates despite costs
Venture capital is backing the bet. Global investments in insurtech rose 19.5% in 2025. In the final quarter, 78% of funding targeted AI-centered investments, up from 42% a year earlier.
The prize justifies the spend. The insurance market is worth $7 trillion. Even small efficiency gains across underwriting and claims processing translate to significant savings.
"Some insurance companies underwrite in the tens of billions of dollars, so every little improvement can really help," Ekstein said.
But the industry's reputation lags other sectors. Insurance ranks lower in public trust than every major consumer industry except social media. That gap creates room for companies using AI for insurance to improve customer experience and build loyalty.
Corgi's growth reflects this opportunity. The company didn't pursue AI to cut headcount. It pursued AI to fix a broken customer experience.
"Right now, pretty much every single business and person in the United States spends about twice as much on insurance per year as they do on software, and the experience is just terrible across the board," Laqua said. "We've focused on using technology to make the customer experience better."
Your membership also unlocks:
