Jamie Phillis faced a surprise $1,200 dental bill after visiting a new dentist closer to home.
AI tools like ChatGPT helped her identify billing issues and submit an appeal to her insurance.
Her appeal was denied, but she never had to pay the bill.
I've always enjoyed good dental health, having just one cavity in my 33 years. Some of this I attribute to having the same dentist since I was 4.
Throughout my adulthood, my dental office felt like the only medical environment where I never had to worry about coverage through my ever-changing insurance providers or getting a massive bill in the mail.
That changed when my well-meaning husband recently suggested I find a dental office closer to our new house. It made sense to forgo a 45-minute commute to my appointments, and I was intrigued by the promises of more modern practices equipped with sleek offices and luxuries like ceiling-mounted TVs.
It was a pleasant experience until I was hit with a massive bill I didn't expect. I turned to AI for help fighting it.
After my first appointment at a new office, I was presented with a $3,320.49 bill
I'd only gone in for a cleaning. To the provider's benefit, I'm an easy upsell, especially for anything health-related. When they recommended an at-home treatment for gum disease, a retainer for my teeth grinding, and a 'full mouth irrigation' for gum health, I agreed, despite them not being covered by insurance and requiring high out-of-pocket costs.
It added up quickly, and before leaving the office, I regretfully paid the $2,111.29 I owed for the items not covered by insurance, which included tax and a credit card fee. I wished I had waited before agreeing to the extra services and done more research before moving forward; however, I naively had no concerns about the remaining balance that would be sent to my insurance company.
A few weeks later, I was floored when my dental insurance informed me that my responsibility was $1,201.80
The primary issue was that, despite finding the dental office on my insurance site, I was paired with the only doctor at the practice who didn't accept my insurance — even after I provided my insurance information both by phone and via an electronic form prior to my appointment.
In my panic, I decided to look over the bill more closely and used ChatGPT to get insights into what I was actually being billed for.
I quickly found multiple instances of bad-faith billing
The first offender was an oral cancer screening test that I didn't consent to, and that isn't recommended or covered by insurance for patients in my age group.
I was also billed for both a periodontal evaluation and a comprehensive oral evaluation — two services that my insurance would not allow to be billed on the same day because they are so similar. Using AI even helped me by flagging that my 'facial photographic images' were charged at seven times my insurance's allowed rate.
ChatGPT also assured me that it's not standard practice to charge a 32-year-old woman $35 for 'oral hygiene instruction' just because the dental assistant reminded me to brush with my toothbrush at a 45-degree angle.
At first, I addressed my grievances the old school way
I called my insurance company, and the representative validated some of my findings but could only direct me to submit an appeal.
I used Otter.ai to transcribe my calls, then uploaded the transcripts of my calls with the insurance company, the receipt from my appointment, my explanation of benefits, and instructions on how to file an appeal into ChatGPT. In under a minute, it had my appeal letter written.
I did go through five revisions — a couple due to me clarifying certain elements of the experience, one because I didn't like how a sentence was worded, and a couple more because I thought of more information I hadn't previously considered including.
Using AI made the whole process feel less daunting
AI was especially helpful when doing the math on things like co-insurance based on allowable amounts. ChatGPT helped me speak the language of insurance companies and convey my thoughts in an industry-appropriate way.
Plus, it formatted everything according to my insurance company's strict instructions and even provided mailing instructions so I could use as little brainpower as possible.
My first appeal was denied
When my insurance company denied my appeal — because the provider scheduled my appointment with an out-of-network dentist — I wasn't happy, but I did feel more confident in dealing with the issue thanks to ChatGPT. I took pictures of the four-page letter and uploaded them into my previous chat conversation.
ChatGPT summarized the letter and suggested I submit a formal second-level appeal and a complaint to the Arizona Department of Insurance. It also advised that I could contact the dental board for my state, as well as my insurance company's grievance department.
I prompted ChatGPT to draft both the second-level appeal and complaint for me, and I submitted them.
After my second-level appeal, my case was never formally resolved, but the dental office never sent me a bill for what wasn't covered. I believe this is because the office had been contacted by my insurance about the billing issues, and they were aware of my appeals and my complaint to the Arizona Department of Insurance.
Ultimately, I never paid the $1,200.
Using AI made the whole process a lot easier
There's a lot of criticism of outsourcing tasks to AI, and I can understand the argument that the instant fixes it provides can make us less equipped to think critically about our problems. However, I think we've all experienced the dread of opening a piece of mail after a long day and finding purposely confusing language that delivers bad news, a big bill, or both.
I don't see the glory in wasting time and mental energy on tasks like researching dental codes or laboring over formatting. I'd rather save my energy for the things I enjoy, or driving the 45 minutes back to my childhood dental office, where I'm now happily receiving dental services again.
When Mirav Steckel opened each of her 15 credit cards, she thought the discounts and rewards they offered would save her money. Instead, her plastic portfolio pushed her into impulse purchases and forced her to reevaluate her spending habits.
"If it was a really bad day at work, my first thought wasn't, 'Oh, let me go home and chill.' It was, 'I'm going to go treat myself to something that's going to make me feel better,' but it got to a point where I was asking people around me for money to pay my cards off," Steckel tells me. "That's not the type of life that I wanted to live."
Steckel chalked it up to financial illiteracy — she was 18, she didn't know the crushing consequences of debt, and signing up for the cards was deceptively easy. Now 21, she has a new trick to follow her monthly budget: use cash. Seeing the physical money rather than making a seamless digital transaction has been instrumental to curbing her spending.
Steckel's frictionless financial experience is an example of the cascade of credit card deals, Buy-Now-Pay-Later plans, and digital personal loans designed to grease the wheels of your spending and make it easy to obscure — or ignore — the purchases you're on the hook for. Consumer spending gurus told me that while these products can benefit those who truly need a financial reprieve, they can also trap people in an endless debt cycle that is difficult to escape. And with the risk of renewed inflation and heightened economic uncertainty, some consumers are more willing to enter payment plans, putting less money down up front in the hope that conditions will improve in the future. The Federal Reserve found that 15% of Americans used BNPL in 2025, up from 10% in 2021. That's coupled with 81% of Americans holding a credit card in 2025.
"If I make you wait or if I make you click through a bunch of things or physically pull out a wallet or, God forbid, cash, that will give you all these moments to pause and rethink things," Scott Rick, a marketing professor at the University of Michigan, tells me. "I just want to be a hot knife through butter here. I want to make it as easy and frictionless as humanly possible."
Welcome to the funny money economy, where credit is king and companies excel at making their products so easy to access that you don't stop to think about whether you can actually afford to pay for them.
'It's gotten so easy to spend'
Maybe you know the feeling: You're online shopping, and you only need that sweatshirt, but when you go to check out, there's a $9.99 shipping fee. If you spend $15 more, though, shipping will be free. You don't really need that jacket that you've been eyeing, but if you buy it, shipping is free. It's a steal. Sure, it's a little more than you'd expected to pay, but if you use a BNPL plan, the first installment fits snugly in your budget, and you can manage the other payments down the line. Congrats, you've been sucked into the funny money economy — fueled by the ease of technology and artful tools to prevent consumers from getting a full grasp on their financial situations.
The modern shopping experience is drastically different from what it was even just a decade ago, Abigail Sussman, a marketing professor at the University of Chicago's Booth School of Business, tells me. It's not only the switch from cash to credit cards, she says, it's also the ability to store your card information on the web and automatically fill it in when prompted. Other forms of payment like Apple Pay make it easy "to spend without even looking at the price, frankly, without even really pausing to internalize it," says Sussman. And even if you consider the number on your screen, it may not be the final final cost.
"I don't think consumers are doing anything wrong. I think it's just that the systems that are in place are really designed to make spending frictionless," she says.
These changes also prey on our brain's desire for instant gratification, says Kristina Durante, a social psychologist at Rutgers' business school. Hundreds of thousands of years ago, humans were solely focused on survival — they weren't thinking about what meal they'd have in two weeks. Our brains are still in that mindset, Durante says. While we can think about the future in an abstract way, the human brain doesn't excel at accounting for the future.
"The part of our brain that wants what it wants when it wants it is so much stronger than the part of our brain that's a brake system that says, 'Wait, hold on, can we really afford this?'" Durante says.
Rick, the University of Michigan professor, says that in addition to the ease of shopping, some companies are using entertainment to pull consumers in. He referenced the TikTok shop, where ads for various products pop up in a user's feed while they're watching videos, and people can purchase them with one click. Or take Disney resorts, where a visitor can tap their wristband or use Disney dollars to make a purchase, leading the consumer to feel like they're using "play money" even as their accounts are slowly drained.
'It's hard to keep track'
The mechanisms that make it so easy to spend also make it easy to get access to debt to spend even more. New technology allows companies to instantly approve new customers for credit cards, and there's an expanded availability of specialized cards, like those intended for people with low credit or ones geared toward students with limited credit histories. These lower barriers are helping fuel a record amount of debt. The latest figures from the New York Federal Reserve show Americans' outstanding credit card balances are at a record high of $1.28 trillion, up nearly 6% from one year ago. Over the same time period, the usage of alternative spending, like Buy Now, Pay Later products, has grown — a December 2025 report from the Consumer Financial Protection Bureau said that the six firms in its sample reported a combined 53.6 million consumers who took out a BNPL loan in 2023, a 12% increase from 2022, at an average amount of $848, up from $725.
When it comes time to settle up your accounts, companies are making it confusing to get a handle on where you stand. Credit cards have variable interest rates that can go up or down depending on a range of factors, including credit score. By offering rewards, points, and limited-time low-interest rates, a consumer might not realize what they're signing up for until it's too late, Ayelet Fishbach, a behavior science professor at the University of Chicago, tells me.
"Points don't feel like much," Fishbach says. "And so if you can convince me that I'm not paying with money, but with some monopoly money such as credit card points, that will obscure the price."
BNPL products similarly use smart marketing and our own mental biases against us. Ying Lei Toh, a senior economist at the Kansas City Federal Reserve, says that while they could help some consumers responsibly make big purchases, they are structured to make people feel "less financially constrained" by breaking the cost into installments.
"It could really worsen the overspending problem and the indebtedness, and the possibility that people would really be spending way beyond their means," Toh says. A CFPB report from 2025 found that Americans are taking on BNPL debt more frequently, and at larger amounts, with the average annual loan increasing from $745 in 2022 to $848 in 2023, up 14%. For credit card companies and BNPL providers, allowing consumers the option to split up payments could generate profit because some afterpay plans come with hefty interest rates and late fees. At the same time, the increased availability of these plans helps companies reach a demographic that might not have traditional credit cards or prefers to make payments in smaller installments.
Stephanie Blanks, 35, managed to escape the BNPL trap — but it wasn't easy. When she had her first child about five years ago, she underestimated how much savings she needed. She ended up maxing out her credit cards and turned to BNPL to buy diapers, groceries, and clothes for her baby. Those small, twice-monthly payments turned into about $3,700 in debt.
"You're like, 'Oh, $10 every two weeks doesn't sound bad at all. I can totally afford that,' until you get 25 or 26 loans in and you're drowning in them," Blanks says. She started paying off that balance at the beginning of September 2025 and hasn't used a BNPL product since. "It was just extremely overwhelming, and I felt like I couldn't get out of the hole," Blanks says.
BNPL plans aren't all bad. When Gabby Raines, 29, moved in with her husband about 10 years ago, they needed a new mattress but couldn't afford it all at once, so they turned to a BNPL plan. "It was a total lifesaver," she says. She has since used BNPL to buy a treadmill and clothes, but even as a savvy, experienced buyer, the ease of signing up sometimes causes her to spend more than she intended.
"We live in a world of such instant gratification and overconsumption, which as Americans we are all guilty of, so sometimes we all should take a step back and really look at the consequences of what we feel like we need right now," Raines said.
The serious consequences of the funny money
While being able to push off or spread out purchases can give you freedom in the moment, the expenses can pile up, and you might find yourself months later with dozens of loans and thousands of dollars in debt that you didn't anticipate.
For some consumers, alternative forms of financing are a means of survival. When the pandemic hit, and she lost her job, Susan Cannon, now 73, used her credit cards to pay for groceries, bills, and complete needed home repairs. While it was necessary, it's also come at a steep cost: Cannon is nearly $40,000 in credit card debt and struggling with sky-high interest rates. "I've always tried to put some in savings, but it's gotten to where it's all going toward interest," Cannon says. "So it's like I cannot get ahead."
For many Americans, though, the funny money economy is a means to fuel nonessential consumption. The Bureau of Economic Analysis's measure of personal expenditures, which includes spending on products like apparel and household appliances, stood at $21.4 billion in the last quarter of 2025, up from $19.2 billion in the last quarter of 2023. The keeping-up-with-the-Joneses culture ingrained in American society has fueled overspending, Sussman says. Your neighbors can only see what you're spending — not the giant debt bill on the backend — which can aggravate the tendency to buy more than you can afford when you're constantly comparing yourself to others.
The rise of social media has made that constant comparison a lot easier, Durante says, because instead of comparing our lives with our neighbors, we can compare ourselves with someone who lives across the country, or the world. "Your brain is categorizing someone far away as someone who is a competitor," Durante says. Social media has also turned spending into a joke, with trends like "girl math," a type of mental gymnastics in which someone might justify using a gift card as free money or consider returning an item as a way to make money. But girl math is really just "human math," Durante says, because it's another way that our brain is thinking about the present without accounting for the future.
"We have a brain built for scarcity living in a world of abundance. And there's going to be a lot of poor decision-making that's made because of that," Durante says.
In the worst cases, funny money can lead to long-term strain. Consumers who fall behind on loans could find themselves facing wage garnishment, and their credit scores could take a hit, making it difficult to rent a home or get an auto loan. Bankruptcy filings increased 11% in 2025, indicating that more Americans are turning to the courts as a last-ditch effort to be absolved of their debt. Higher debt loads slow economic growth, as consumers put less money into the economy and more toward paying back what they owe.
The very nature of funny money, though, is that these tough consequences can be under wraps until they come crashing down all at once. And it's unlikely these trends will reverse anytime soon, given pervasive economic uncertainty, Fishbach says. High inflation might lead a consumer to think that it makes economic sense to purchase something now even if they cannot fully afford it, "because the price might be higher later, tariffs might make it higher," Fishbach says. And the existence of mechanisms like debt forgiveness, bankruptcy, and Buy Now, Pay Later products all feed into the funny money mindset, where it's nearly impossible to predict your financial situation a year or even a month from now, making it easier for consumers to put off payments.
"People are smart, but they are busy, and you are not required to get a master's degree in economics before you go to the grocery store," Fishbach says. "We created a system that makes it very hard to make good financial decisions. "
Ayelet Sheffeyis a senior reporter on Business Insider's economy team, covering education, student loans, and the federal workforce.
Kinetic services cars attached with sensors that will be critical for an autonomous future.
Courtesy Kinetic
Modern cars are equipped with sensors to support driver assistance systems and safety features.
Kinetic is a startup that provides sensor calibration after a car is involved in a collision.
Kinetic CEO Nikhil Naikal said his company aims to service autonomous vehicle fleets.
Human drivers aren't the only ones who need to have eyes on the road.
Many cars on the road today are equipped with at least half a dozen sensors, from cameras to radars, to support safety features and advanced driver-assistance systems (ADAS) that are now ubiquitous in the modern automotive industry. And as automakers continue to offload the task of driving from humans, cars will become even more sensorized.
Kinetic, a Southern California-based startup, wants to scale a service that will fix those sensors if a car ever gets into a collision. The CEO analogizes it to the modern car's optometrist.
"We have eyes, and when we need to correct vision, we go to an optometrist who places all these letters at 20 feet, measures our vision and prescription, and then gives us the glasses to correct our defect," CEO Nikhil Naikal told Business Insider. "In the same way, this is a digital prescription to correct the errors of the car's understanding of the world around it."
No car is the same as it once was after its first fender bender.
Panels get bent, parts are replaced, and the paint won't match the original exactly.
The same principle goes for the sensors on a car, Naikal said. Sensors are highly sensitive to alignment, and a fraction of a degree can significantly affect the effectiveness of an ADAS or self-driving feature.
Once a vehicle is in a front-end collision or rear-ended, those sensors could be thrown out of position. Mechanics can put the sensors back in place, but they won't be re-mounted in a spot perfectly identical to their factory placement.
That's where Kinetic's robotics platform and software come in to calibrate the sensors — or, as Naikal put it, give them a "digital prescription."
The auto shop of the future
An 8,000-square-foot facility in San Francisco's Dogpatch neighborhood is one of eight Kinetic hubs on the West Coast.
The inside is unlike loud, messy auto body shops. A Kinetic facility is quiet and mostly empty save for a rotating platform and a robotic arm attached to a short track. A typical location is staffed with up to two technicians, Naikal said.
On a Wednesday afternoon, Kinetic was servicing a 2022 Toyota Camry Hybrid LE that had been repaired at a local body shop after a front-end collision.
Kinetic uses robotics to calibrate sensors on a car, post-repair.
Lloyd Lee/BI
The car was driven onto the platform, where Kinetic's camera package takes detailed photos of the car, and a robotic arm points a laser at the radar sensor embedded in the Toyota's grill.
Kinetic's software then puts the car and the sensors back in sync.
"What the car thinks it's going to do and what the sensors think the car is going to do needs to be aligned," Naikal said. "The sensor thinks that the car is going to go right, whereas the car actually needs to go left — that's a problem. It causes unstable things like ghost braking and jerking."
The entire process takes about 10 minutes. Naikal said a single Kinetic hub can service about 80 cars a day.
Some repair shops might rely on a mobile service, in which a specialized technician comes to calibrate the sensors. Naikal said a lot of body shops don't have the space, lighting, equipment, or trained technicians that can accommodate an in-house service. Local body shops can either choose to send cars to Kinetic's hub or, if they have space, lease the company's equipment.
By the end of 2026, Naikal said he aims to have 20 hubs in the US.
Kinetic's platform can also perform damage inspections and develop repair plans, but Naikal envisions servicing autonomous vehicle fleets, which require constant cleaning and sensor maintenance, as another line of business. The future, he said, will require a new kind of service infrastructure built around robotics and software rather than a traditional body shop.
"It's going to be more than just Jiffy Lubes and Valvolines," Naikal said. "We think of ourselves as the infrastructure layer for the future of autonomy."
Alibaba.com president Kuo Zhang says AI is aiding the surge in one-person companies.
OpenClaw, an AI agent, has helped boost the trend in China.
Alibaba.com faces tariff-related challenges. Zhang says he focuses on two things to navigate them.
China has seen an explosion in "one-person companies" thanks to a little help from AI agents.
Alibaba.com president Kuo Zhang told Business Insider he has seen this growth firsthand, estimating that 30% to 40% of the e-commerce platform's customers are "solo entrepreneurs."
Business Insider previously reported on the rise of one-person companies, or OPCs, which rely on AI tools such as agents and vibe-coding technology to build their businesses without hiring other employees. Some Chinese cities are attracting these startups by offering free housing, rent-free offices, and subsidies of up to $720,000.
Zhang said AI agents have helped make these types of startups possible. He runs Alibaba.com, the Chinese cloud and retail giant's e-commerce platform that connects buyers and suppliers. While most of its customers are from China, it's also growing its customer base in the US, Europe, Latin America, Southeast Asia, and other regions.
One-person companies face barriers to building their business. Tasks like uploading products to different sites, managing social accounts, and handling customer complaints are "easy for a big company" but harder for small businesses, Zhang said. While these tasks may not be the expertise of small businesses, they are "essential for success," he added.
Agents can help with some of that grunt work.
"Instead of taking the place of the human beings, actually, they are the employees of that solo entrepreneur," Zhang said of AI agents.
Alibaba.com recently launched Accio Work, an AI agent designed for small businesses, including one-person companies. It can help companies manage daily e-commerce operations, including customer service, tax compliance, marketing, logistics, product listings, and more.
"They are in lack of help or tax support. And now, AI is very easy to use. It's very easy to adopt and to understand everything is going to change that perspective, and we think they can benefit from them the most," Zhang said.
Alibaba.com's main Accio agent first launched in late 2024 and now has 10 million active users a month, according to the company.
China's OpenClaw craze
The rise of OPCs has been boosted by OpenClaw, the open-source AI agent that has become wildly popular in China. The craze has sparked a sort of OpenClaw gold rush and spawned quirky agents for stock trading and setting up blind dates. Many OPCs have also built businesses using the OpenClaw agent.
Zhang says that the rise in OpenClaw has helped educate the market about AI agents. Alibaba itself introduced JVS Claw, a mobile app to help users install and deploy OpenClaw more easily.
Compared to China, Zhang said American users are less educated about OpenClaw and that there's a smaller scene around it. That said, OpenClaw has issues with security and return on investment, he said, adding that some customers have spent "hundreds of US dollars for tokens," and when they don't get the results they want from using agents, they quit.
Instead, it's important for agents to be user-friendly, secure, and easy for customers to get started, he said.
"If you go to SMBs, you ask about all the fancy terms about AI, like the token economy, like cloud, like open cloud, probably they've never heard about that, and what they care most is about how these tools can help me," Zhang said.
Alibaba faces changing tariff policies
Alibaba.com and businesses on the e-commerce platform face another major challenge: navigating a constantly changing tariff landscape. Zhang first became president of Alibaba.com in 2017, during President Donald Trump's first term. As policies change, Zhang said it helps to focus on two things: serving customers based on supply and demand, and technology.
"Technology changes on a daily basis, so I've learn a lot from the tech companies, and from our SMBs. They will tell us what they need and how we can leverage tech the best to help them on a daily basis," Zhang said.
"The rest, I say, is noise," he added. "Just follow all the rules, and we follow all the rules in the world."
"Informed" Polymarket users made $143 million since 2024, according to a new study.
Researchers analyzed most of Polymarket's trades between 2024 and 2026.
One co-author told BI he hoped the analysis would inform prediction-market regulation.
Last June, a new Polymarket account with the name "ricosuave666" began betting thousands of dollars on specific questions regarding Israeli military strikes on Iran.
The gamble paid off. When Israel struck Iran in the early hours of June 13, ricosuave666 pocketed roughly $155,000 before going dormant for seven months. The account resurfaced in January 2026 to place new wagers — before analysts flagged its activity and it was promptly deleted.
Ricosuave666's moves were among more than 210,000 suspicious Polymarket trades analyzed by researchers at Columbia Law School and the University of Haifa. The trades netted $143 million for the "informed" traders who made them, according to the researchers, whose findings were published this month.
The most suspicious trade made by ricosuave666 was only the 3,662nd-most unusual trade in the data, authors Joshua Mitts and Moran Ofir wrote. They said ricosuave666's gains closely matched ill-gotten gains attributed to a person charged by Israeli authorities with using military secrets to trade.
The study, which analyzed most of Polymarket's trades between 2024 and 2026, is the first to provide an estimate for the total amount won by suspicious accounts, whose activities are often flagged by market observers and traders on X and Discord chats.
The authors used five criteria relating to trade timing and amounts wagered to screen for accounts that made big, bullish bets shortly before news broke, though they acknowledged that their methods could have been over-inclusive or under-inclusive.
"We don't have any reason to think that the unobserved relationships are cutting in one way or another," Mitts, a Columbia Law School professor who has previously written about potential insider trading in equities markets, told Business Insider.
Mitts told Business Insider that he and Ofir generally used the term "informed" trading rather than "insider" trading because some of the biggest trades they flagged took place in markets where too many people influence the outcome for it to be rigged, like those related to Donald Trump's 2024 election. "Informed" is a broader term that encompasses trades made by smart bettors as well as those with unfair advantages.
The study said the election bets were included with other "anomalous" bets because the authors didn't want to be accused of massaging their methodology to get a particular outcome.
"We think there's going to be a lot of regulatory attention. We see this as just the beginning of the conversation," Mitts said.
No proof of insider trading
The authors said it's possible that they flagged profitable trades that were part of a hedging strategy and offset by a loss in another wallet controlled by the same user. Still, they characterized the volume of suspicious trades they identified as a "conservative lower-bound estimate of anomalous profits."
Harry Crane, a Rutgers statistics professor who has written about prediction markets and was not involved in the study, questioned its methodology. While the study ranked over 210,000 bets based on how unusual the bets were for the traders who made them and for the markets where they were made, the ranking of their "suspiciousness" was heavily skewed by profitability, rather than what ordinary people would consider suspicious, he told Business Insider.
"A winning bet is getting disproportionately more weight than a losing bet of the exact same type," Crane said.
Most of the 20 most suspicious trades identified in the paper, accounting for about $16 million of the $143 million in profits reaped by flagged accounts, related to the 2024 election results. Others related to Federal Reserve decisions and sports match-ups, where manipulation is theoretically possible or insider information theoretically could have leaked.
Mitts said the plan was eventually to publish all the data used for the study.
"Prediction markets have outpaced the legal frameworks designed to govern them," the authors wrote. "Our paper aims to provide the empirical grounding and legal analysis necessary to close that gap."
Polymarket changed its policy
Prediction markets have been around for decades, initially as academic experiments and later as a tool popular with members of the so-called rationalist community. The general idea is highly accurate guesses about the future can emerge from a group of non-experts who put their money where their mouths are.
Some advocates for prediction markets see insiders cashing in as a feature, not a bug. Shayne Coplan, Polymarket's founder, said last year that it's "cool" that Polymarket "creates this financial incentive to divulge information to the market."
Earlier this month, the company announced that it was banning trades by people with "stolen confidential information" and "illegal tips" as well as trades by people who can influence the outcomes of events they're betting on.
It's not clear how the company will enforce the prohibition when it doesn't know who its users are. While the company has a regulated US subsidiary, that entity processes less than 10% of the volume of trades as the offshore exchange, which doesn't collect users' names or other identifying information beyond an email address.
It didn't reply to requests for comment from Business Insider.
Kalshi announced earlier this year that it is seeking fines from two users who broke its rules, including a video editor for MrBeast who placed bets on what words would be said on shows before they were released.
Prediction markets grew slowly for years. In 2025, Kalshi grew rapidly after it began offering Americans the chance to bet on sports in a way it calls more fair than sportsbooks. Problem gambling experts are concerned that prediction markets pose similar risks to sportsbooks, however, and several states have sued Kalshi and other prediction markets, arguing that they amount to unlicensed casinos.
The Commodity Futures Trading Commission, a federal regulator, fined Polymarket in 2022 and generally prevented prediction markets from offering many contracts until Trump took office last year.
Michael Selig, a Trump appointee who took the helm of the regulator last year, has been a vociferous advocate for such markets, and criticized the states for their crackdowns.
Spencer Rascoff talked "founder mode," swipe fatigue, and Gen Z dating habits with Business Insider.
Bloomberg/Getty Images
In just over a year, Spencer Rascoff has radically transformed how Tinder operates.
Rascoff changed up the org chart, pushed back on narratives of fatigue, and recentered on what Gen Z wants.
He told Business Insider about his leadership — and why a long-married man would want to run a dating company.
Spencer Rascoff met his wife before Tinder. And eHarmony. And Match.com.
He was 17 years old, attending a barbecue for students who planned to attend Harvard College. They got to Cambridge, started dating, and have been together since. His next love-match was startups, founding a slew of tech and media companies, including his opus, Zillow.
Then came Match Group, the conglomerate that owns apps like Tinder and Hinge. He knows what you're thinking: What qualifies this long-married man to know what singles want? He's never had a Tinder hookup gone wrong; he's never longed for those in Hinge's rose jail.
"I am living proof of how important it is to find the right person," he told me at Match Group's sunny West Hollywood office. "Were it not for her, it would be awful."
Rascoff thinks you can find that same life partner on his apps. The market seems to disagree, with stocks down across the category in recent years. In the post-pandemic dating app boom, Match Group's stock traded around $150; now, it hovers around $30. Five days before I spoke with Rascoff, the S&P 500 announced that it would expel Match Group.
Rascoff needs to make online dating sexy again. He's done it before — think of how stuffy the homebuying process was before Zillow. But Zillow was a disruptor, where Match Group is a legacy player. It'll take bigger swings to get more Americans swiping right.
Inside Rascoff's Tinder redesign
Rascoff doesn't come across like the stereotypical "founder mode" type. He's calm and quiet-spoken, often stopping mid-sentence to consider exactly what he wants to say.
Power centralized around Rascoff soon after he joined Match Group. He took over as CEO of Tinder from Faye Iosotaluno. Two of Match Group's longest-standing faces have also since left: Hinge founder Justin McLeod and chief operating officer Hesam Hosseini. The COO position will not be refilled.
Rascoff quickly reformed Tinder into a series of independent pods. He subscribes to Amazon's two-pizza rule: a team should never be so big that they'd need more than two pizzas. He also made the company flatter. It put Tinder in line with how he ran Zillow and how Hinge was already running, he said.
"Tinder product and engineering used to be a very large, monolithic organization where the priorities of what gets built came from on high," he said. The org chart change "unleashed an enormous amount of innovation that was buried."
He also drilled into the app's mission statement: "Tinder is the most fun way to spark something new with someone new." The team now says it in unison before every company meeting, something Rascoff said he values, even if it can be "awkward" and "creepy."
Spencer Rascoff led the keynote at Tinder Sparks.
Henry Chandonnet/Business Insider
Multiple Tinder leaders told me that the dating app now runs more like a startup. (Two leaders directly used the term "founder mode.") It's an ironic twist, given that Tinder itself was born in a conglomerate: IAC's Hatch Labs. But Rascoff seems bent on ripping up any remnants of bureaucracy.
Right before the team went on stage at the Tinder Sparks conference, Rascoff approached Claire Watanabe, Tinder's vice president of product. He was using the new Music Mode, and wanted to know why he couldn't hear it on his profile, she said.
"He's into the details," Watanabe said. "He has an opinion."
Navigating a sour dating app market
Rascoff and I mostly abstained from talking about Hinge. Part of that is because of the event: We're here for Tinder Sparks, a moment when Rascoff can play Steve Jobs to talk about a litany of new features.
Hinge is "on a path to be a billion-dollar business," Rascoff said. "The way they've done that is by knowing whom they're building for, and by bringing consumer insights into what they're building."
Tinder, on the other hand, has slumped. Its annual downloads have been shrinking since 2023, according to Appfigures. In 2021, 61 million people downloaded Tinder, the firm found. In 2025, that figure dropped to 48 million downloaders. Meanwhile, Appfigures found that Hinge was consistently showing positive growth in annual downloads.
Make no mistake: Hinge maintains a fraction of Tinder's user base, and Tinder is still the biggest dating app in the world. But the warning signs are certainly flashing as red as a Tinder flame.
The Tinder Sparks event was a moment to turn it all around. Rascoff spoke confidently on stage, looking casual in his blue jeans, and outlined a vision for the future of dating. We sipped oat milk lattes and smiled for the photo booth. It's hard not to be excited about an app when it's named in a Bad Bunny song.
The Tinder Sparks event included product announcements, specialty juices, and a bracelet-making bar.
Henry Chandonnet/Business Insider
The room's optimism didn't seem to match Tinder's perception in the outside world, where we hear constantly about "swipe fatigue" and the decline of online dating. Women especially report low-quality matches; a 2022 Pew study found that more women reported negative than positive online dating experiences.
When I asked Rascoff about fatigue, he chose his words carefully. "Some people have left the category because they find dating apps tedious," he said. His goal is to introduce "fun" features to combat that feeling.
M Science research analyst Chandler Willison said that some investors have begun to think that those worries — about "systemic issues" and "industry-wide malaise" — weren't as unchangeable as they once believed.
"Spencer has done a really good job pushing back against that idea," Willison said.
A new generation of daters
Rascoff may be married, but he's still swiping.
He excitedly showed me his phone. His Tinder profile had a photo of him, his wife, and his dogs. His bio says that he's "just here for research about our product."
It's not his first go-around on the apps. A few years after his father died, Rascoff's mother got on the apps. He served as his mother's "dating copilot," helping her improve her profile and respond to messages.
For Tinder to bounce back, he'll have to aim his dating advice down a generation. Rascoff talks about Gen Z constantly. He went directly from Match Group's Gen Z employee resource group to our interview, he said. He cracks jokes about Gen Zers who bring up astrology in their job interviews, and shows off a Rubik's Cube with Tinder's imaginary Gen Z customers.
Tinder imagines Gen Z daters like "TK" and "TK"
Henry Chandonnet/Business Insider
Match Group isn't the first company to prioritize Gen Z. It's why Nespresso brought in Dua Lipa, and why the Duolingo owl raved to "brat." For Match Group, though, the generation is life-or-death. You might still need espresso or a language app when you're 45 and happily married, but you won't need Tinder.
I'm a Gen Zer myself, a 23-year-old Tinder swiper. It was an odd experience to sit in the Tinder Sparks auditorium and be told by older generations how I want to date. I want to be "perceived authentically," they said. I'm interested in "self-development," they said.
During our interview, Rascoff reminded me that Gen Z wants "lower pressure" ways to date. That's why they pivoted to in-person events; because they're chill, and because they're "meeting Gen Z users where they are."
The low-pressure model made sense. I'm not dating for marriage; I just want to have a good time. But why, then, are all my friends on Hinge, the higher-pressure alternative? (Hinge lets you like fewer people, and doesn't provide the anonymity of a two-way match.) I could name at least four Gen Z couples in my life who met on the app designed to be deleted.
For a while, all I heard about Tinder was that it was an endless hole of one-night stands. Then, I met up with a friend, who told me that she'd started dating. She let me in on a secret: it wasn't Hinge she was using, but Tinder. It was easier, she said.
There's little real data to back up a Tinder turnaround yet. Monthly downloads remain at about 3.9 million and haven't started climbing, per Appfigures. Willison said that Tinder was still in a "recovery" phase. But my friend's confession felt like a sign. People can (and do) still find sparks on the app.
Rascoff said he thinks he'd do well on Tinder. Those who are authentic and put in the hours find their spark, he said. "If I were single, I would certainly do that."
Tech investor Ron Conway, seen here in 2013, is hoping that California Gov. Gavin Newsom helps kill a proposed wealth tax before it reaches voters.
Jeff Chiu/AP
Tech investor Ron Conway is worried that California's proposed wealth tax could pass.
He and several other big names in tech are funding campaigns to stop the proposed ballot initiative.
Conway said it's imperative that the proposed tax never reach the ballot.
Famed Silicon Valley investor Ron Conway says he wants to kill California's proposed wealth tax now.
"Our job is to get Gavin to negotiate this so that it doesn't get to the ballot. So, maybe they don't get the signatures," Conway told Jack Altman during an episode of Altman's "Untapped" podcast that was posted on Wednesday.
Conway, the founder of SV Angel and known as "The Godfather of Silicon Valley," said that if the proposed wealth tax reaches the ballot, it "could" pass.
A recent UC Berkeley Citrin Center for Public Opinion Research-Politico poll found that support was hovering around 50%, within the margin of error of potential failure, though it is still very early in the process.
Major names like Google cofounders Sergey Brin and Larry Page have already rushed to move assets out of California, the state home to the most billionaires. If passed, California residents with a net worth of over $1.1 billion would face a one-time tax totalling 5% of their assets. Supporters of the initiative are still gathering signatures ahead of a June deadline.
Conway said Gov. Gavin Newsom, who is publicly opposed to the wealth tax initiative, is aligned with the efforts. Conway said one way to give Newsom bargaining power is by supporting the three competing ballot initiatives, which would effectively neuter the proposed wealth tax.
Brin, Stripe cofounder Patrick Collison, former Google CEO Eric Schmidt, and others have poured over $44 million into "Building a Better California," a political action committee that is pushing the three competing anti-wealth tax ballot measures. In November, Conway donated $100,000 to "Stop the Squeeze," another group that is opposed to the proposed tax.
Venture capitalist Marc Andreessen once called Conway "the human router," a nickname he told Altman that he considers a compliment. Conway has been a fixture in tech for decades, making early bets on Google, Facebook, and other companies. OpenAI CEO Sam Altman, Jack's brother, credited Conway with helping him hold OpenAI together during his brief ouster in 2023.
Conway also told Jack Altman that their interview could not run late, because that night he had courtside seats at the Golden State Warriors game next to House Speaker Emerita Nancy Pelosi and her husband Paul.
"We must keep this off the ballot," Conway said. "So a whole bunch of work has to happen for that."
