AI in Insurance
Most carriers are silent on AI - and that silence has a cost
Clients deploying artificial intelligence are almost certainly insured for it. The problem is that nobody, not the carrier, not the client, and often not the broker, can say with confidence which policy covers what.
That ambiguity is the central finding of a 2026 RAND Corporation report, The Insurability of Artificial Intelligence, authored by researchers Sasha Romanosky and Celine Robinson. Drawing on public AI incident data, US litigation records, state legislation, and admitted-market insurance filings, the report maps a market split three ways: a minority of carriers affirmatively covering AI losses, a growing number filing broad exclusions, and the majority remaining silent. It is that silent majority which creates the most immediate problem.
Silent coverage is not actual coverage
When a policy says nothing about AI, coverage depends on how a claim is characterized at the time of loss. MORE
News
Treasury Secretary says Federal Insurance Office should stay out of state-based regulation
Treasury Secretary Scott Bessent praised state regulation of insurance Tuesday at a House Financial Services Committee hearing.
He agreed with a committee member, Rep. Troy Downing, R-Mont., that the U.S. Treasury Department's Federal Insurance Office should not encroach on the state-based insurance regulatory system.
The Federal Insurance Office "is an observer," Bessent said. "We facilitate communication. We do not exercise authority."
Downing, a former Montana insurance commissioner, asked Bessent, "Do you believe that Treasury should minimize its insurance activities, to clearly define federal responsibilities and leave insurance regulation to the states?"
"One hundred percent," Bessent said.
Warren Buffett steps down as chairman of Berkshire Hathaway
Warren Buffett is stepping down as chairman of Berkshire Hathaway in the next step of the transition at the top of the $1 trillion conglomerate he led for more than six decades.
Buffett, 96, will become chairman emeritus and will take on his new role immediately, the company said Friday. His oldest son, Howard, will become Berkshire Hathaway’s new chairman, part of a long-planned succession. Howard Buffett has been on the Berkshire Hathaway board since 1993.
Warren Buffett, known by a legion of financial market followers as the Oracle of Omaha, will remain a director. He had been Berkshire Hathaway’s chairman since 1970, but he took control of the struggling New England textile company in 1965.
Greg Abel took over for Buffett as the CEO of the $1 trillion conglomerate at the start of this year though Buffett continued to come into the office each day to look for new investments and deals and offer Abel any advice. It’s not clear whether Buffett will change his routine at all now.
Financial Results
Progressive posts weakest combined ratio of 2026 as agent book growth continues
Progressive reported a combined ratio of 89.3 for August, a 6.2-point deterioration from 83.1 in the same month last year and the sharpest year-over-year gap the carrier has posted in 2026. Net income fell 22% to $951 million from $1.22 billion in August 2025, following July's 12% decline. Two consecutive months of year-over-year net income drops, with the gap widening each month.
Net premiums written reached $7.61 billion, up 6% from $7.20 billion a year earlier. Total policies in force grew 7% to 40.5 million. Direct auto led growth at 9% to 16.88 million policies; agency auto grew 7% to 11.34 million. Commercial lines added a more modest 4% to reach 1.25 million policies.
Research
An insurer failure is often followed by several others
Some 40 insurance companies failed in 17 countries in 2025.
The Property and Casualty Insurance Compensation Corporation (PACICC) has just published the fourth edition of its summary of insurance company failures around the world. Insurer failures occur worldwide, and 2025 was a “particularly challenging” year, according to Alister Campbell, CEO of PACICC.
The authors of the Global Failed Insurer Catalogue 2026 are Grant Kelly, chief economist at PACICC, and Judy (Zhe) Peng, research associate. In this research, the authors report 1,273 failures in 98 countries since 2000, in both property and casualty insurance and life insurance.
In the July 2026 issue of the quarterly Solvency Matters newsletter, Campbell noted that in 2025, some 40 insurance companies failed in 17 countries. “The large number of insolvencies last year represents both a sudden upward jump in an otherwise downward-sloping trend over the past five years, and a result even higher than the 25-year average of 35 insurer failures annually,” Campbell wrote.
“Our research identifies clusters of insurer failures that occurred 118 times across 59 jurisdictions since 2000, often after sustained periods of relative calm. This should serve as a sobering reminder to all financial services sector stakeholders of the risk of complacency,” he adds. PACICC made the same recommendation when it published the third edition of the catalogue in 2025.
InsurTech/M&A/Finance💰/Collaboration
Insurtech company Luzern Risk raises $45m in Series B round
The company said the new funding will go towards developing its AI-native technology platform.
Luzern Risk, a US-based full-service captive insurance manager focused on alternative risk solutions, has secured $45m in a Series B funding round.
Insight Partners led the round, with Trust Ventures and existing backer Caffeinated Capital also taking part.
Caffeinated Capital previously led Luzern Risk's seed funding round in 2023 and its $12m Series A round in 2025.
The company said the new funding will go towards developing its AI-native technology platform, which is designed to cut the time required to launch and manage custom captive insurance programmes at scale.
The proceeds will also be used to support the systemisation of operations to reduce turnaround times and broaden client options across the alternative risk value chain.
Luzern Risk CEO and co-founder Gabriel Weiss said: "We set out to make captives more accessible to a broader set of the market, and everything we have learned since has strengthened our conviction that captives will play a far bigger role in risk management than they do today."
Orion180 Insurance valued at $1.14 billion as shares fall in debut | Reuters
Orion180 Insurance (OIG.O), was valued at $1.14 billion after its shares fell in its Nasdaq debut on Friday, setting a cautious tone for other insurers eyeing public listings later this year.
Shares opened at $11.50 after the Melbourne, Florida-based company priced its IPO at $12 per share, below its marketed range of $15 to $17, raising $240 million. Jumpstart your morning with the latest legal news delivered straight to your inbox from The Daily Docket newsletter.
The debut comes amid a difficult start to the fall IPO season, as concerns over AI spending, the Federal Reserve's recent interest-rate hike and rising bond yields weigh on investor appetite for new listings.
Kenneth Gregg, founder and CEO of Orion180, however, said, "It was the right decision for our business and our team regardless of macro environment."
State News
California moves to ban auto insurance surcharges for single drivers
The proposal would eliminate a 30-year-old practice where unmarried drivers pay up to $100 more for identical coverage compared to married policyholders.
California Insurance Commissioner Ricardo Lara announced proposed regulations to prohibit auto insurers from using marital status as an optional rating factor.
If finalized, the rule will eliminate a 30-year-old practice where unmarried drivers pay up to $100 more for identical coverage than married policyholders.
The proposed regulatory action follows a July 16 appellate court decision in Ison v. Lara confirming the commissioner's legal authority under Proposition 103.
Lawyer for State Farm fined over AI hallucination in Los Angeles lawsuit
Sept 14 (Reuters) - A California judge has ordered a lawyer for insurance giant State Farm to pay a $999.99 fine after finding that some of her filings included nonexistent cases stemming from her use of an artificial intelligence program.
Here are the details:
- Judge Elizabeth Bradley of Los Angeles Superior Court said in her order on Friday that submissions from the attorney, Jacquelene Robinson of law firm Musick, Peeler & Garrett, included seven case citations that did not exist.
- The underlying lawsuit, filed in 2024, involves insurance coverage for fire and storm damage at a house in Carson, California. State Farm is a defendant and has denied any wrongdoing.
- Lawyers for the plaintiff said they discovered the citation errors. “The issue is pervasive and present in nearly every single motion in limine filed by State Farm in this case,” Eric Khodadian, a lawyer for the plaintiff, told the court.
- Robinson did not immediately respond to a request for comment.
- In a court filing, Robinson apologized and said she took full responsibility. “I recognize that citation to fabricated authorities, especially those included in submissions to the court, is unacceptable,” Robinson said.
- State Farm in a statement on Monday said it “expects all external counsel to adhere to the highest standards of professionalism and ethical conduct, including when using AI and when reviewing and confirming the accuracy of all court filings.”
- State and federal judges have sanctioned lawyers in dozens of cases when an AI program made up citations, misquoted the law or, in at least one case, invented witness testimony.
New Jersey bill targets AI-powered insurance claim denials
A New Jersey bill would bar insurers from using artificial intelligence to make the final call on claim denials.
Assembly Bill 5494, introduced on September 14, 2026, by Assemblyman Chris Tully of Bergen County, covers three personal lines - homeowner, automobile, and flood insurance. The bill's statement points to "inaccuracy, unfair discrimination, data vulnerability, and lack of transparency" as the risks AI poses to policyholders.
The restriction is narrow. A5494 would not pull AI out of claims handling entirely. Insurers could still use automated tools for triage, assessment, fraud scoring, and approvals. What the bill would prohibit is letting an algorithm deliver the final "no" - the denial itself. The bill's statement describes this as "requiring a claims adjuster or investigator to review" denied claims, though the operative text frames the rule as a prohibition on AI making that last decision rather than an affirmative mandate for human review.
Fraud
Insurance Fraud Costs $45 Billion per Year, and Honest Policyholders Often Pick Up the Tab | Insurify
Fraud, inaccurate information, and misrepresentation cost the U.S. personal auto and homeowners insurance markets an estimated $45.3 billion annually, according to an analysis released by insurance risk-assessment company Clearspeed.
The analysis estimates the cost of inaccurate information and fraud at $31.6 billion annually for auto insurance and another $13.7 billion for homeowners insurance.
And those losses don't necessarily stay on insurers' balance sheets.
Instead, insurers can account for expected fraud and other losses when setting rates, passing on at least some of the expense to policyholders through premium increases.
The National Insurance Crime Bureau (NICB), citing research from the Coalition Against Insurance Fraud, estimates that insurance fraud results in roughly $900 in additional annual premium expense per policyholder.
The FBI has also estimated the impact at $400–$700 annually for the average U.S. family.
"It's embedded in loss costs, goes through the rate-making process, and ultimately ends up being a pretty important factor in what the premium levels are," Rob Hoyt, a risk management and insurance professor at the University of Georgia, told P&C Specialist.