Inside the Research That Connected Litigation Frequency and Litigation Funding

The Demotech Difference had the opportunity to speak with Joseph Petrelli and Todd Kozikowski about the significance of their collaboration on a research project of immense importance to those focused on the increasing frequency of litigation, social inflation, and nuclear verdicts.

Demotech’s research project was initiated following the failure of several insurance companies, despite securing decades of affirmations from actuaries and independent auditors. Subsequent internal due diligence revealed that, even with favorable actuarial opinions and unqualified, “clean,” independent audits, these insurers experienced escalating levels of annual litigation that, over the four-year period prior to failure, grew to six times their market share of premium or policies in force.

Recognizing that general economic conditions in the affected jurisdictions had not changed during that period, and suspecting that only technology or “something new” could drive litigation growth at that scale, Petrelli turned to Kozikowski, an established data scientist and technologist with experience in the insurance sector. The goal was to analyze, detect, and explain whether this “something new” caused the explosive growth in litigation, and how it remained largely undetected despite more than 120+ actuarial opinions on loss and LAE reserves and more than 120+ independent audits in the years prior to the failures.

What follows is a conversation that explores how that investigation began, what it uncovered and how those insights can help insurers, regulators, and other stakeholders better understand emerging litigation risk.

Early Years and Influences

The Demotech Difference (TDD): Thank you for taking time to speak with us. Todd, tell us a bit about yourself, where you were raised.

Todd Kozikowski (TK): I was born in western Massachusetts and raised in Rhode Island and outside Boston. I’ve been a New Englander my entire life. My father was a first-generation college graduate, and before that, our family history was rooted in military service and hard, manual work. Relatives worked demanding jobs, from commercial fishing and tending looms in the mills of Fall River to door-to-door sales for Electrolux.

My grandfathers and uncles were part of America’s Greatest Generation. They returned from World War II with stories that stayed with me and shaped my values around hard work, ethics, and treating everyone with kindness and respect.

I was also fortunate to grow up surrounded by an incredible group of mentors through an astronomy club. From an early age, I was drawn to the stars and planets, building telescopes, and exploring the universe. They taught me science and engineering, but more importantly, they instilled a deep sense of curiosity about the world around us, both here on Earth and far beyond the treetops.

Joe Petrelli (JP): For me, it was Rochester, New York. My dad and all his brothers were house painters. They were high school dropouts. They served in the Korean War. My mom was the first in her family (and our family) to graduate from high school. After graduating, her father asked her, “What college are you going to?” Instead, she earned her beautician’s license and opened a salon. By 1949, she had three shops. She sold them in 1951 because she was pregnant with me. The sales price was $2,900.

When she went back to work, she was on the assembly line at Delco, in a time when people stood, and the line moved. Workers picked up a piece, applied the next step, and put it back down. Mom received suggestion awards for ideas that improved production, usually a U.S. government savings bond for $25. Before the end of each fiscal year, her supervisors would say, “You’ve already gotten all the awards we can issue, Annie.” She would say, “Okay, I get it,” and keep giving suggestions anyway. With her, it was never about the money. It was about getting the job done properly and as quickly as possible.

Lessons from Family and Community

TDD: Todd, how did your upbringing shape your perspective on people and relationships?

TK: The main lesson was that everyone pulled their weight and helped each other. You learned to respect everyone, and you understood that any person might be the one who shows up when you need help. Teamwork and accountability became second nature, whether on a fishing boat, in a mill, or within a family.

One story really stayed with me. My great uncle, who I never met, was a baker during the Depression. I only heard about him through stories. When my great-grandmother was in her 80’s, she would give me money and ask me to go to the corner store for milk and eggs.

I remember one day at the store, I bought a few items and went to the register. The woman at the counter looked at me and asked, “Are you Francisco’s grandson?” I said, “Yes. How did you know?” She said, “I could tell. You’ve got the family eyebrows.” When I asked how much we owed, she said, “No charge.”
I asked why. She said, “Your great-grandfather fed our entire family during the Depression. I was a little girl. He had nothing, but he gave food to families in the area until he went out of business. This is my way of returning the favor.”

That story taught me that actions matter, and they echo for generations. The idea of paying it forward has stayed with me, and I try to live by it.

Education and Early Career

TDD: Todd, you attended Bates College. Tell us about your education. Is there a course or a professor who shaped your views on the importance of assisting others?

TK: It was transformative and humbling. I majored in physics and mathematics and was surrounded by very strong critical thinkers. My thesis advisor was a demanding and highly respected theoretical physicist whose classes reshaped how I learned to think.

In one of my advisor’s upper-level theoretical physics classes, our first exam had only a few questions. I struggled through all of them and was sure I had done poorly. When we got the exams back, I had scored a seven. It turned out seven was about the class average. She admitted she might have made the test too hard, but what she was really doing was teaching us how to think independently and then as a team. That emphasis on rigorous thinking and collaboration has stayed with me.

I also spent time in nuclear physics labs at MIT working around a reactor in Cambridge. Those experiences reinforced the same lessons: take time to think deeply on your own, then bring your ideas back to a broader group to test them. It shaped how I approach complex problems and how I rely on teamwork for the best outcomes.

TDD: Before we get too far into your current career, what was your first job?

TK: My first real job was working for the dean of surgery at Harvard Medical School. At the time, I thought I wanted to be a doctor. He hired me to support clinical trials and research in nutritional medicine. He invented gastric bypass surgery and helped coordinate the original food pyramid. He was a former Navy medical doctor and ran a tight ship.
He specifically wanted someone with a math background rather than a traditional epidemiologist. He believed someone in physics and mathematics would ask different questions and look at data differently. I learned a tremendous amount about healthcare and medicine, but I also realized that clinical medicine wasn’t the right fit for me. That experience pushed me toward technology and data early in my career.

JP: My first job was a morning paper route and an evening paper route. You learn a lot about people. I received tips, thanks, and gifts on holidays. If someone wanted their paper at the back door, you put it at the back door. Otherwise, I tossed the folded paper on the porch. I collected weekly: some paid weekly, others paid every two weeks, others once a month. I don’t recall being stiffed, but I do recall learning the importance of cash flow.

Connecting with Demotech

TDD: Todd, how did you first meet or become acquainted with Joe and Sharon Petrelli and Demotech?

TK: I met Joe around 2014 or 2015, when I was with a start-up software company that focused on big data and analytics. That’s where our paths first crossed. Years later, we reconnected when Demotech launched this research project.

JP: I remember the meeting with that start-up company. When we went out to its website, Todd showed me how one could peel back the veneer of the website and view personal information on the visitor to the website based on IP address.

The “Something New” Behind Carrier Failures

TDD: On the research project: Joe emails you. He’s getting hammered on these insolvencies and tells you, “There is something new going on.” Did you think he was looking for cover on the carrier failures?

TK: No. My first reaction was that Joe has always been interested in answering the “why” behind a problem, not just the “what.” So, I assumed there was a significant issue that needed attention.

At the time, I was working on other projects, but when Joe described what he believed had happened in Florida, that these insurers had effectively been litigated to death, it got my attention. We all see the TV commercials and billboards for law firms. You can’t drive through a city without seeing them. But I also knew from my earlier work that the real story might be what was happening online.

I had spent years working with marketing technology and early analytics platforms, including a company called Unica that came out of MIT. We were looking at segmentation, digital behavior, and the early building blocks of SEO and online intelligence. So, when Joe described what he was seeing, I suspected that search, digital targeting, and technology might be driving the change.

When I started digging into the data in the spring of 2022, the first thing that jumped out was the online activity of a public adjuster that was bidding on insurance-related keywords and spending more than $650,000 a year. They were specifically targeting insurance companies, including mobile homeowners in Louisiana. You don’t spend that kind of money unless the business model is working.

That was the moment I went back to Joe and said, “There’s a business model here.” From there, the picture expanded. It wasn’t just public adjusters. We saw plaintiff firms actively targeting insurers, along with other opportunists. It became clear this wasn’t just about property insurance or Florida. The same digital patterns were showing up in other lines and other states. This was a national issue.

We also knew there had to be a response. These carriers were being targeted in ways that traditional tools weren’t designed to see. We wanted to help companies build a kind of digital moat around their business and, at the same time, lessen the impact on policyholders, who ultimately pay the price when litigation costs drive up premiums.

Why Todd Was the First and Only Call

TDD: And, Joe, when you discuss this, you always say Todd was your first and only call. Why do you say that?

JP: Todd has been a technologist and data scientist with experience in insurance. The work he had done was innovative, and he had experience in research. Given the meteoric growth in annual litigation at each failed carrier, I believed technology had to be part of the “something new.” Actuaries and auditors had been signing off on year-end financial statements for decades, despite the unique and well-documented market conditions in Florida and other jurisdictions. Finding my “something new” required someone who had been on the front lines of analyzing data and performing research where none had previously been done.

I also knew Todd to be someone I could trust. If the “something new” was harmful to stakeholders, he would want to respond in a way that helped. Demotech’s internal research showed that litigated claims grew dramatically at these carriers, yet that reality went largely unnoticed despite more than 120+ “clean” audits and more than 120 statements of actuarial opinion on loss and LAE reserves in the years leading up to insolvency. An explanation was needed. Once a cause was identified, a solution needed to be created.

TK: When I went into the work, I tried to assume nothing beyond the fact that something meaningful was happening. The research quickly showed orchestrated digital targeting.
As I met with P&C carriers of different sizes and tested our findings, it became clear that this was not just a claims issue or an SIU issue. It was also a marketing and legal issue. A cross-department response was needed, and the tools to support that response didn’t really exist.

Once a claim is filed, you can analyze it, but the damage has already started. The larger question is how to prevent the next claim, how to stop opportunistic or fraudulent claims from forming in the first place. That became our focus: helping companies see upstream, understand how they were being targeted, and prioritize the investigations that would make the biggest difference.

This became especially important because SIU teams and regulators are overwhelmed. They are getting hit with more activity than they can reasonably process. If you can help them see where the heaviest and most coordinated targeting is coming from, you can help them focus.

When Joe spoke to Congress about these findings in November 2023, most people in the room had never heard of this business model. That tells you how hidden it was, even though its impact was actually very visible when looking at loss trends.

Uncovering the Online Business Model

JP: We are discussing this topic in late 2025 in preparation for the February 2026 edition of The Demotech Difference. What amazes me is that Todd found this online business model within a week. He was reporting on targeting: “Joe, It’s online. It’s digital. It’s SEO. It’s pay-per-click. It’s cloaking. There are all these tactics.”

With these insights, Demotech started our own “keyword” research in the fall of 2022. Based on articles we found online, the business model was using litigation platforms and early forms of artificial intelligence around 2015-2017.

Executing our research, Todd put together the pieces of the puzzle of what destroyed the failed carriers. He identified the business model of tech-enabled litigation: search engine optimization, online tactics, litigation platforms, and litigation marketing companies. Once you know those tools are in play, your online research of the problem is different. Demotech, Todd, and I unearthed this business model and, now, each facet and mutation is a piece of the puzzle.
Todd recognized early that this was not just a Florida problem. It was countrywide. Todd and I looked beyond “nuclear verdicts” and “social inflation” to identify a proximate cause for the surge in litigation. The attraction of third-party litigation funding is due to the success of the online business model.

TK: One thing we kept hearing when we shared these findings with carriers was, “If it’s a claim, it’s too late.” The root question is how policyholders are being solicited and intercepted in the first place.

The targeting we uncovered turns potential claimants into clients of law firms or other intermediaries before the insurer even has a chance to talk to them. That means every claim starts from a more adversarial place. Carriers lose the conversation with their own policyholders because a third party got there first using digital tools.

Litigation Platforms, ABSs, and Third-party Funding

JP: When you go back to specific things that we found, subsequently, an article in the National Law Journal dated Nov. 15, 2021, stated that in 2015, Reuven Moskowitz approached John Morgan to develop a litigation platform to facilitate litigation. They built Litify by 2017. Four years later, Morgan & Morgan reported a 500 percent increase in revenues due to a 500 percent increase in caseload.

Knowing what emerged from the research project and Todd’s findings, we re-examined the creation of alternate business structures (ABSs) in Utah, District of Columbia, and Arizona. In the Commonwealth of Puerto Rico, alternate business structures became an opportunity on Jan. 1, 2026. The development of ABSs required revision or modification to American Bar Association Rule 5.4(d). Revisions permit direct ownership, investment, or management of law firms by non-lawyers. ABSs eliminate the frictional cost of segregating otherwise unpermitted concurrent practices.

Investing in litigation accelerated markedly between 2015 and 2020. The legal profession saw the opportunity to invest in litigation platforms to enhance revenues, scale their model, and attract third-party investment. Third-party investors wanted in because the online business model was remarkably profitable.
The business model is lucrative. That’s why the legal profession proposed changes to ABA Rule 5.4(d) and began creating ABSs. The opportunists have a multi-year headstart on the sectors they are targeting. With an industry focus on fraud, it was Todd and I who identified opportunism as a lever increasing litigation frequency.

Building a Team and Responding to the Risk

TK: Once we understood the scope of the problem, it was clear this couldn’t be a one-person effort. I pulled together people I had worked with before who understood data, analytics, and how to look for signals in big, messy datasets.

We started with search manipulation and ad campaigns, then quickly saw how this was evolving into AI-generated content and coordinated digital behavior. People were injecting misleading content into large language models, creating deepfake sites, and automating outreach. Attacks were becoming faster, broader, and more sophisticated.

Being early to detect this risk gave us an advantage. We could see the patterns forming and help companies intervene before the targeting became claims and litigation.

We often talk about social inflation as if it just “happens.” In my view, what we’re seeing now is closer to a social inflection point, driven by the success of this tech-enabled litigation business model. Third-party litigation funding is growing because the underlying model is working.

You have funding, AI-driven platforms that help decide where and how to file cases, and a legal environment that can produce nuclear verdicts and class actions. The opportunists realized that technology could help them find and convert potential claimants at scale, often by intercepting basic online searches.

Florida as a Testing Ground

JP: What better place to incubate this online business model than a state that (once) had 8 percent of the homeowner claims in the country, yet 75 percent of the homeowner’s insurance litigation? I speculate that the business model was developed and perfected in Florida because the existing level of litigation provided camouflage and smoke cover for the online business model. What better place to hide?

The insurers that failed had a 3 percent market share of premium. Yet, in the year they were destroyed, they had a near 20 percent market share of new litigated claims. When you have a constant 3 percent market share, and your litigated claims rise over four years to near 20 percent, there is “something new” going on.

How the Business Model Has Evolved

TDD: Todd, when you first undertook Demotech’s research project, it was April 2022. For the February 2026 edition of The Demotech Difference, nearly four years later, has this business model evolved?

TK: It has moved from a focus on search engine optimization and digital ad buying into full-blown AI manipulation.

We’ve identified nation-state and cyber-criminal organizations for hire on the dark web. A subset of opportunists is hiring these groups to target carriers in the United States online. By outsourcing to those groups, the opportunists distance themselves from the mechanics of the attack.

We also see the funding side. One recent publication identified 35 confirmed litigation funders in the U.S. At 4WARN, we have detected 125, and we believe the actual number is several times higher.

The money is not simply going to help plaintiffs bring meritorious claims. It’s being used to grow the business model, by paying for marketing agencies, building brands, and funding additional targeting. The attack vectors are numerous, but once you can measure them, you can prioritize them.

To help companies do that, we developed the 4WARN Risk Score™. It gives a structured benchmark using four main categories:

Who is targeting you and how? Law firms, adjusters, contractors, third-party groups, or criminal organizations.

How strong and coordinated is the targeting? How much money is behind it? Is it one bad actor or a network?

How are your policyholders being solicited? What tactics are being used to turn basic inquiries into adversarial claims?

What are your own vulnerabilities? Your tech stack, endpoints, and current preparedness.

Traditional cyber security focuses on what happens behind the firewall. We focus on the external digital ecosystem, where opportunists are using search and content to manufacture new forms of risk.

AI as a Fraud Vector

TDD: Todd, you’ve talked about the use of AI. For most people, it’s a term they hear, not something they see. From what you’re seeing online, how much of a headstart do the bad actors have on AI?

TK: AI makes it easier and faster to create content and automate outreach, which means it can accelerate fraud. The growth curve is steep.

Most people think of AI as tools like ChatGPT. Many organizations think about AI primarily in terms of productivity. At 4WARN, we look at AI as a fraud vector. We see several major use cases:

Content injection: inserting misleading narratives into large language models so that future queries resurface that bias.

Deep fakes and lookalike websites: sites that mimic legitimate brands and capture personal information.

“Group-think” among AIs: one model picks up a false narrative embedded in it, others reference it as fact, and suddenly multiple systems are reinforcing the same untrue story.

Automation of claims and outreach: bots that can generate or escalate claims at scale.

What we call “Ghost GPT”: cracked or retrained models that are deliberately optimized to commit fraud. These are being sold on the dark web.

This gives people who cannot code, and who may not deeply understand the technology, the ability to launch complex attacks. They can simply buy the tools. We also see organized criminal groups using these techniques, then reinvesting the profits into litigation funding, and more targeting.

A Real-World Example

TDD: Do you have an example of this type of activity?

TK: In December, I spoke at a long-term care conference at a national carrier’s headquarters. I discussed new fraud vectors in long-term care and how online tools are being used to target families trying to find care for their parents.

I talked about how our phones, and tablets are constantly listening for signals and how we’ve all had the experience of casually mentioning a destination like, “I’d love to go to Aruba,” and then seeing ads for Aruba shortly afterward. I explained persistent data and what that means for targeting.

About 30 minutes after my presentation, an attendee from an SIU team came up to me and said, “You’re not going to believe this. Since you started speaking, I received a text and a voicemail about nursing home facilities and care for my mother. I’m not actively looking. The only place this was mentioned was in your session. My phone must have been listening.”

That’s a small example, but it shows how quickly signals can be picked up and used. These attack vectors are real, and they are getting more sophisticated.
You asked about the greatest online risk facing organizations in the near future. I think the biggest risk is the belief that these attacks are inevitable and cannot be prevented. AI and automation can imitate legitimacy at scale. But prevention is possible if you are monitoring the right signals. It’s like having a burglar alarm. You are grateful that the alarm goes off as soon as someone opens the door.

How 4WARN Maneuvers the Internet To Alert and Protect Those Being Targeted

TDD: Tell us a little bit about how you maneuver the internet to assist companies being targeted.

TK: 4WARN looks at the world from the outside-in. We scan the internet and public marketing ecosystem; nothing is installed on a client’s systems. We don’t need their internal data to tell them how they are being targeted. We bring the targeting intelligence to them.

Today we synthesize three quarters of a trillion data points as signals of risk across search engines and marketing platforms. We actively monitor 20,000 opportunists across 1,500 companies every day and watch for emerging patterns.

That brings me back to Joe’s decision to send that first email. If he hadn’t reached out, I believe many more carriers and policyholders would have been harmed before anyone understood what was happening. The work we have done with departments of insurance and regulators has helped them prioritize their investigations. Our clients have strengthened their websites and marketing, adjusted underwriting, and in some cases negotiated better reinsurance because they can quantify and reduce this type of risk.

If that research project had never started, some of those companies might still be out there, heavily targeted, without realizing why their claims and legal costs were spiraling.

The Surprise of What Was Discovered

TDD: Todd, you’ve been a technologist and data scientist for decades. When Joe said there was “something new” but didn’t know exactly what it was, did you think you might find what you actually uncovered?

TK: No. I expected to find some form of digital acceleration, but I didn’t anticipate the level of systematic, coordinated, and dark activity that we eventually uncovered. The networks, the tools, and the sophistication of some of these actors went well beyond what I imagined at the outset.

Third-Party Litigation Funding and What Comes Next

TDD: Joe, there is an effort countrywide to discuss disclosure of third-party litigation funding. Do you think that will stem the tide in the level of litigation or the amount of litigation funding targeting insurers and other brands online?

JP: No. Disclosure requirements will be used, and likely already are being used, to restructure investments so they need not be disclosed. Tech-enabled claim instigation is a holistic and ever-evolving business model. The first mutation was alternate business structures. The next mutation is managed services organizations. Law firms pay taxes. Insurers pay taxes. Insurance buyers pay taxes. Federal taxation of net profit from the business model would be a step in the right direction.

TDD:   Todd, as the person who discovered the “something new” that Joe first sensed, what are your thoughts on the journey since day one?
 
TK: The online business model had been operating covertly for years, possibly even a decade.  When several insurers in Florida failed in rapid succession, Joe needed to understand why then. He identified dramatic growth in litigation counts during the four years preceding liquidation, despite market conditions remaining largely unchanged.  

That’s when he called me.  Together, we developed a research project to determine whether there truly was “something new.” Through that work, I identified the tech-enabled business model that had contributed to the collapse of those carriers. 
 
Joe asked me to help understand what had changed. Once we did, the focus shifted to how organizations could defend themselves against similar targeting. That work ultimately led to the creation of 4WARN, a service designed to help organizations detect and respond to coordinated digital risk before it escalates.
 
TDD:   Joe, as the first to suspect that “something new” was affecting the insurance industry and contributed to these liquidations, we’ll give you the last word. 
 
JP: Proponents of litigation funding and tech-enabled litigation instigation have been attacking us since our research discovered their online business model.
TDD: Todd, time for a lightning round of personal questions. Favorite meal?
TK: Pizza.

TDD: Favorite snack?
TK: Chips and salsa.

TDD: Favorite dessert?
TK: Apple pie or carrot cake.

TDD: A book you’d recommend and why.
TK: Given my astronomy background, one of my favorite books is “Starlight Nights.” It was written by an astronomer from Delphos, Ohio. I love it because of his passion and his contribution to the field. I also appreciate that amateur astronomers can make real contributions from a telescope in their backyard.

TDD: Favorite movie?
TK: “The Godfather.”

TDD: A talent that few people know you have?
TK: I play piano. I have competed in and finished nine Ironman races.

TDD: Motto, credo, or slogan you live by?
TK: Nothing less than 150 percent.

TDD: A hobby that you have or plan to develop?
TK: I love oil painting.

TDD: Dog or cat person?
TK: Dog.

TDD: Joe, your turn — your favorite meal?
JP: Salmon on a Caesar salad.

TDD: Favorite snack?
JP: Day-old pizza.

TDD: Favorite dessert?
JP: Carrot cake or apple pie.

TDD: A book you’d recommend and why.
JP: “Principles of Marketing” by Philip Kotler. This book was used in the marketing course when I went back to school for my MBA. Because of this course, I decided to research title insurance. As a result, in 1992, Demotech was the first company to review and rate title underwriters.

TDD: Favorite movie?
JP: “Blazing Saddles.” An oldie but it was one of the silliest movies ever made. Sacrilegious about everything. When I saw it, there was one person in the audience who laughed at every joke three seconds after everyone else had stopped laughing. I’ve never had a theater experience like that again.

TDD: A talent that few people know you have?
JP: Although I rarely grin or smile, I have a healthy sense of humor.

TDD: Motto, credo, or slogan you live by?
JP: There is no traffic jam on the extra mile. Also,
know competition, no monopoly. No competition,
know monopoly.
TDD: A hobby that you have or plan to develop?
JP: If it requires hand or eye coordination, I am not your person. However, I look forward to being an active volunteer for my church.

TDD: Dog or cat person?
JP: Dog.

TDD: Any closing comments?
JP: Demotech’s internal analysis, and our research project that led to Todd’s findings, unearthed the previously covert, online efforts capable of accelerating litigation at industrial scale by targeting specific carriers and entire jurisdictions. Bringing those dynamics into the open has allowed insurers, regulators, and other stakeholders to better understand the “something new” that they face so that they can respond more thoughtfully.

TDD: Thank you both for speaking with us today.