This man believes auto insurance companies will soon use AI to track driving quality and reimagine the underwriting process, prompting viewers to question how likely this outcome may be.  

On Dec. 16, TikToker Matt Layson (@matthewlayson) posted a video theorizing that drivers will soon see an entirely new “subset of auto insurance” that uses an app to consistently track the driving patterns of drivers in exchange for “much better” pricing. 

Layson believes insurance companies will use this day-to-day data to influence future terms and pricing. 

How Does Layson Believe That AI Will Alter The Auto-Insurance Industry? 

Key to Layson’s speculation is that insurance companies will use AI to completely re-imagine how these companies charge their customers. 

“The more information you feed an AI, the more accurate and the better it gets at predicting what’s gonna happen,” he says. “So can you imagine an AI looking at constant driver logs and updates for hundreds of thousands of drivers?”

He believes that AI will use predictive modeling to determine which drivers are more likely to have accidents in the next 3-5 years and use that information to “charge more for those drivers” or even “non-renew them.”

He clarifies that he is very nervous about this possibility. 

“I don’t know how to feel about that. I can see a very easy, direct correlation of it being a way more cost-effective auto insurance kind of program,” he said. 

However, he doesn’t fully trust the insurance companies implementing AI like this. 

“But I could also very easily see the insurance company using it for evil. Being like, ‘oh well, we can predict all sorts of other things, and we can charge based off of that.’”

Do Any Car Insurance Companies Already Use AI?

The idea of using AI in auto insurance is already happening, but mostly behind the scenes and in clearly defined ways. 

Insurers are increasingly using machine learning and AI models to analyze telematics data, fine-tune underwriting, and reduce claims costs, with a large majority planning to expand AI investments in the coming years. These systems already improve risk prediction, personalize premiums, and automate claim handling far more than traditional methods.

Many commenters were quick to point out that many insurance programs already track the driving of their customers. 

“Pretty sure State Farm already does this,” wrote one comment. “I get a decent discount for having their app on my phone that monitors my driving.”

“Mine already does that. I have a tracker in my car that records a lot of data like speed, location and g-forces. It gives you a score for every drive,” another shared. 

“Progressive already has this for their snapshot – all licensed drivers. The policies have to have the app to be tracked for the snapshot discount to apply or to get lower rates for good driving. I have opted out, no thanks,” a third commented.

Layson responded to the previous comment about Progressive, clarifying that he is theorizing something separate from these already-existing programs. 

“That program has been around for years,” he wrote. “What I’m talking about doesn’t exist yet because of legal problems with an AI black box spitting out predictions without a direct mathematically proven explanation of why an insurer wants to change a price for some aspect of the risk.”

How Is Layson’s Theory Different From Current Programs Offered By Insurance Companies That Monitor Driving Patterns And Use AI?

While companies like State Farm and Progressive have programs that ask users to install an app or plug-in device to track basic driving habits and reward safer behavior with discounts, this differs from what Layson is specifically addressing. 

These programs collect data on mileage, speed, braking, and acceleration to generate a driving score. This infomation is used to adjust premiums or provide a discount for relatively safe drivers. “Pay-how-you-drive” and “pay-as-you-drive” plans are already widespread, and many drivers see lower rates when they opt in because those plans analyze actual behavior rather than relying solely on age and ZIP code.

What Layson is warning viewers about is a world where that continuous data stream isn’t just rewarding good driving, but becomes a core, AI-driven part of underwriting itself. Instead of optional “discount” plans, he imagines systems that continuously predict a driver’s future accident risk years out and price or even refuse coverage based on that prediction. 

That’s a shift from current usage-based models tied to voluntary telematics toward a far more proactive, automated, and opaque system. He predicts that an AI “black box” would ingest months or years of behavior and make discrete choices around pricing and renewal without traditional actuarial transparency. Which is not how today’s telematics programs are regulated or explained to consumers.

How Feasible Is Layson’s Speculation? Are There Any Legal Barriers Preventing Such?

Layson’s theory resonated with viewers, prompting commenters to express fear about the possible repercussions. “Rates won’t go down,” read one comment. “Their profits will increase. No one will save.”

“I will not,” expressed an additional commenter. “I will pay more before I let the insurance have all my data. It would have to be outrageously more expensive.”

Where Layson’s scenario becomes more speculative, and much less likely in the near term. This is in the deployment of continuously streaming, fully automated predictive models that supersede traditional rate filings and actuarial disclosure. While insurers can use AI to assess risk, current implementations must fit within existing regulatory frameworks. This require transparency, fairness, and justification for pricing decisions. 

In the U.S., state regulators require rate filings, and some states, like California, have strict limits or bans on telematics-based rating altogether, treating the tracking of behavior data as a special case requiring disclosure and consent. 

Additionally, regulators are already issuing guidance around the responsible use of AI in underwriting to help prevent unfair discrimination and require explainability in pricing models. The National Association of Insurance Commissioners (NAIC) and state departments of insurance have published principles and rules that would constrain completely opaque “black box” AI from being used without oversight. These steps act as legal and practical barriers to the exact scenario Layson fears, at least without significant regulatory change.

While carriers are moving toward richer data and more advanced predictive modeling, the leap to always-on AI that arbitrarily reclassifies drivers’ risk years in advance would represent a significant departure from how auto insurance pricing is legally governed today.

MotorBiscuit requested statements from Layson via TikTok direct message. Additionally we have contacted Progressive and State Farm via email.

@matthewlayson

Here I am predicting the future, ever in motion it is. But I know insurance companies are moving into AI more and more, I know insurance companies really like telematics, and they want to expand the usage of that. I know insurance companies are talking more about making their driver telematics programs mandatory and that they can surcharge clients based off of driving habits. How else do you see AI changing auto insurance?

♬ original sound – Matt Layson