Insurance startup uses AI to provide coverage in high-risk California and Florida

A San Francisco-based insurance company using artificial intelligence is promising to reform home insurance in parts of the US facing some of the worst climate risks.

In 2024, a group of technology veterans from Silicon Valley started Stand insuranceusing what the company describes as physics-based AI models to underwrite policies and create customized fire resilience plans for properties in California’s wildfire-ravaged areas, where homeowners often struggle to get coverage from traditional carriers.

Last month, Stand raised $35 billion in a Series B funding round to expand coverage in hurricane-ravaged Florida, where homeowners have struggled to find insurance options from a limited number of carriers and are facing skyrocketing premiums.

“Traditional ways of pricing insurance don’t work in an environment that is so unpredictable,” Stand co-founder and CEO Then Preston told earlier The Wall Street Journalreferring to climate-related risks.

Realtor.com® has contacted Stand for comment.

Stand currently provides coverage for California homes valued between $3 million and $15 million, with a combined total insured value of more than $1 billion statewide.

What sets Stand apart from typical insurance companies is what it describes as its science-based approach to risk assessment and mitigation.

“We use AI to understand the true risk of your property and help you make your home insurable,” according to Stand’s website.

Most standard insurers determine whether or not to insure a home largely based on its location: if a property is in a fire or flood zone, the company is more likely to deny coverage or increase premiums to protect the bottom line.

Stand, on the other hand, takes into account not only the location of the property, but also the efforts of homeowners to make their homes more resilient.

Here’s how it works: Stand’s software creates what the company calls a “digital twin” of each home, based on data about the materials used in its construction — and even the type of tree species planted in the backyard.

AI models then simulate how fire, wind, embers and water would affect a property in the event of a disaster, assigning a quantified percentage to each risk.

The next step is to develop a customized plan to reduce the risks and make the property safer, such as replacing the windows or shingles, clearing the brush around the house or swapping a wooden fence for a steel fence.

California’s insurance crisis

Fire risk, mitigation and insurance have been at the center of the debate in California this year, following January’s deadly wildfires in Los Angeles County, which caused economic losses between $76 billion and $131 billion, with insured losses estimated at $45 billion, according to the UCLA Anderson Forecast last updated in March.

Across the Golden State, roughly 1.3 million homes with a combined rebuilding cost of more than $791 billion face a moderate or greater risk of wildfire, according to a U.S. government study. recent report by Cotalitya real estate data analytics company.

Nationally, almost 3 million homes are threatened by forest fires.

Even before the devastating fires in Palisades and Eaton, California was already in the grip of an insurance crisis, with homeowners in at-risk areas struggling to get coverage at all costs.

After several large-scale infernos swept through California in recent years, some of the largest private insurance companies have stopped offering new policies in high-risk areas, and those that remain have been allowed by the state to raise rates.

The withdrawal of private insurers has forced hundreds of thousands of homeowners to seek coverage through the HONEST planCalifornia’s last resort. As of June, the state-backed insurance company reported total exposure of $650 billion, up 42% since September 2024.

The number of policies on FAIR Plan’s books increased to more than 610,000, an increase of more than 30% compared to nine months earlier.

Florida is the next frontier for AI

Against this challenging backdrop, Stand looks to offer wealthy California homeowners an alternative as it prepares to do the same in Florida, providing hurricane and wind protection to homes and condos priced from $500,000 to $2 million.

A Cotality’s 2025 Hurricane Risk Report found that Florida is the state most at risk for hurricane and wind damage, with more than 8.1 homes exposed to moderate or greater risks.

Repairing or rebuilding these properties would cost more than $2.3 billion.

Like California, Florida has seen insurers balk at the increased risk levels and withdraw from some of the most vulnerable areas, in part because of the high costs of litigation.

Lately, however, there have been signs of improvement, with more airlines offering policies in Florida and rate increases slowing thanks to legal reforms and regulatory changes.

As a result of this shift, Florida’s insurer of last resort Citizens Property Insurance Corp.which has served as the state’s largest airline, has reduced its policy burden as part of what it calls its “depopulation program.”

By the end of the year, the number of Citizens is expected to shrink by more than 510,000 policies, which will be taken over by private insurers. WUSF.

As of early November, Citizens still had about 560,000 policies, down from a peak of 1.4 million two years ago.

Industry experts have their say

However, industry experts and stakeholders are divided on whether it is a good idea to take the human element out of insurance.

“I think using AI is a great tool, [but] it is not a solution for acceptance,” Jamie Levenshonsenior vice president of the commercial real estate division at Insurance Agency of America and board member of the Florida International University School of Real Estatetells Realtor.com. “You lose the human element in acceptance, which I think is quite dangerous”

Levenshon claims that every homeowner has a different level of risk tolerance, and she argues that insurance professionals can come up with “much more creative ways” to structure a comprehensive coverage plan than an AI model could.

Levenshon doesn’t dismiss AI modeling out of hand, but says it has limitations, especially in Florida, where the risks and mitigation techniques differ greatly from those in California.

“Wind and water work very differently than fire,” she explains. “The mitigation that has to come from a roof being blown off is very different than the mitigation from smoke. … During a wind event, a hurricane, the wind blows for eight hours. It’s a completely different animal.”

She adds: “I think there is a place for AI in underwriting to the extent that it can better model the risks, but as far as the underwriting element of it, I’m skeptical because I think you still need humans to evaluate risk.”

Critics have also pointed out that AI models tend to be siled and opaque, making it difficult to verify their accuracy, or to question their risk assessments in the event that a homeowner’s claim is denied or the monthly premium goes up, because Bloomberg reported.

Lindsey Klarkowskipolicy vice president for data science at the National Association of Mutual Insurance Cos., says that using AI in insurance has several benefits for consumers, including identifying ways to make homes more resilient, fraud prevention and detection, and increased efficiency.

“It’s important to note, however, that using AI is just one part of the equation, especially in the context of coverage availability and pricing,” she tells Realtor.com. “AI can lead to greater precision in risk-based pricing and has proven effective in certain mitigation efforts, both of which can lead to cost savings. But those cost savings could be offset by massive losses due to changing weather patterns, litigation excesses and abuse, or otherwise, underscoring the need to focus on cost drivers that the industry can control.”

Asked about the possibility of claim denials based on AI ratings, the NAIMC official said that in Florida, existing state insurance law still applies.

“Artificial intelligence is a tool that is added or incorporated into processes that are already within the scope of existing insurance law,” Klarkowski explains. “It is not an end goal for the insurance companies around a state’s statutes, regulations and consumer protections.”

She adds that new technological developments in insurance will overall benefit consumers and the wider market.

“Whether it is better understanding and more accurate pricing of risk, improved mitigation, or operational and other efficiencies, AI use could potentially address challenges for Florida, as it does for other markets and homeowners across the country,” Klarkowski concludes.