The Telematics Bargain Was a Lie: 69% of Drivers Aren't Saving Money — and Regulators Are Finally Paying Attention

Usage-based insurance was supposed to be the fairest deal in the history of auto insurance: drive safely, pay less. Ten years and 21 million enrollees later, the Maryland Insurance Administration has released data that exposes the bargain as something else entirely. In the most comprehensive state-level analysis of telematics outcomes, only 31.2% of enrolled drivers saw any premium decrease. Meanwhile, 23.6% paid more — with some increases reaching 42.5% of total premium — and 45% saw no change at all. The Consumer Federation of America drew the obvious conclusion: "The supposed savings from telematics are a mirage for many drivers, or are at least highly exaggerated." Insurers are still running ads promising "up to 40%" discounts. The gap between those two realities is where this industry's credibility problem lives.

The 31% Problem: How Insurers Buried the Real Telematics Math

The marketing math has always been straightforward: enroll, drive safely, save money. The actuarial math is far less generous. Maryland's 2025 survey of 16 insurers covering 303,845 enrolled drivers found that the modal outcome of telematics participation is no change whatsoever. Less than a third of participants benefit from the programs they were sold as universally beneficial. The MIA also received over 811 consumer complaints in 18 months specifically related to telematics, representing 12.4% of all private passenger auto complaints — a disproportionate share for programs covering roughly 13% of the state's drivers.

Consumer surveys complicate the picture further without resolving it. AutoInsurance.com's survey of 1,282 telematics users found 66% experienced some decrease — but median savings among that group were $27 per month, nowhere near the advertised maximums. Consumer Reports' far larger survey of 40,566 policyholders found a median annual savings of $120. Both figures are positive but modest, and neither accounts for the 59% of respondents who said they'd reconsider enrolling if they knew their premium could rise.

The structural problem is how these programs are designed. Insurers that allow premium increases — Allstate, GEICO, Liberty Mutual, Progressive, and Travelers — have financial incentive to use behavioral data not just to reward safe drivers but to re-price risky ones. The savings for the 31% who benefit are partly financed by the surcharges levied on the 24% who don't. The framing of "discount program" obscures what is, functionally, a behavioral surcharge program with a marketing exemption.

Data Acquisition With a Discount Sticker: What Telematics Was Always Really For

The savings narrative was always the least interesting thing about telematics from an insurer's perspective. The real value is risk segmentation and data monetization at scale. Progressive's Snapshot program now covers 45% of new auto business, and industry veterans like Robin Harbage, a Progressive telematics veteran, have been direct about the mechanism: well over half of drivers cause far fewer than half of all accidents, making telematics most powerful as a tool for identifying and retaining preferred-risk drivers — not for rewarding the average enrollee.

Allstate took this logic to its extreme. Through its data analytics subsidiary Arity, Allstate built what the Texas Attorney General called "the world's largest driving behavior database" — over 45 million Americans' driving data, collected at a rate of more than one billion miles per day. The collection method wasn't Snapshot; it was a hidden SDK embedded in third-party apps including Life360, GasBuddy, and MyRadar, paying those developers to silently track users. Arity then sold that data not just to insurers but to retailers and marketing companies, functioning as, in the Texas AG's framing, "part software developer, part data broker, part AdTech company."

LexisNexis built a parallel infrastructure through its Telematics Exchange, ingesting driving data from automakers including Kia, Mitsubishi, and Subaru, then reselling it to insurers. The company was developing a lead generation product that would give insurers "direct access to prospective customers who meet your target parameters based on near real-time driving data." John Davisson of the Electronic Privacy Information Center described the product plainly: "There's no earthly justification for a third party to vacuum up behavioral driving data and sell it off as sales leads."

This is the business model that the "save up to 40%" messaging was built on top of.

Surveillance Creep: When Driving Scores Become Lifestyle Scores

The Maryland survey documented that insurers are collecting more than 50 data points per driver, including location and GPS routes, time of day, phone usage, vehicle diagnostic codes, turn signal usage, and adaptive high beam engagement. Location data is particularly consequential: insurers including Allstate, GEICO, and USAA track destinations, which means risk algorithms can effectively penalize drivers for frequenting neighborhoods classified as high-risk — a well-documented proxy for race and income.

Time-of-day scoring is the clearest example of how behavioral pricing embeds structural discrimination. Almost every insurer penalizes driving between midnight and 4 a.m. Low-wage workers — overnight hospital staff, warehouse workers, delivery drivers — have no choice about when they work. Research on telematics discrimination has noted that Black and Latino workers are overrepresented in overnight shift work, meaning a pricing factor ostensibly based on individual behavior functions as a demographic surcharge. The New York Department of Financial Services warned in 2024 that external consumer data and AI systems must be monitored for "unfair and unlawful discrimination" — a warning that most telematics architectures are not currently designed to address.

Only 38% of telematics users believe their insurer's data collection is always accurate, according to J.D. Power's 2025 Insurance Shopping Study. That confidence gap matters because inaccurate data can trigger premium increases with no meaningful appeal process — a structural injustice Maryland's proposed SB 351 legislation would begin to address by granting consumers the right to correct data errors and requiring insurer governance plans.

The Regulatory Awakening: What Enforcement Actions Actually Reveal

The FTC's action against GM and OnStar, finalized in January 2026, is the most significant signal of what regulatory attention to this space looks like when it arrives. The settlement bars GM from sharing consumer data with consumer reporting agencies for five years and requires affirmative customer consent for data collection for 20 years — a sweeping corrective to what the FTC found was a "misleading enrollment process" that funneled precise driving data to LexisNexis and Verisk without informed consent.

Texas went further in January 2025, suing Allstate and Arity in the first-ever state enforcement action under a comprehensive data privacy law. The Texas case makes explicit what regulators have long suspected: that telematics infrastructure operates as a shadow data brokerage, subject to none of the disclosure norms consumers expect from their insurers. Verisk, facing its own exposure after a 2024 New York Times investigation, quietly shut down its driving behavior data history product rather than defend it.

At least 15 states have introduced bills targeting telematics and connected-vehicle data since early 2024. Maryland's SB 351, backed by Insurance Commissioner Marie Grant, would require insurers to notify consumers when telematics data triggers a premium increase, establish data correction rights, and authorize the Maryland Insurance Administration to limit what data types can be used for pricing. A prior version passed the state Senate in 2025 before stalling in the House — a near-miss that reflects how quickly the political ground is shifting.

Why Opacity Is the Industry's Biggest Liability

The core problem for insurers is not that they collect data. Behavioral pricing is actuarially sound and, in principle, more equitable than crude demographic proxies. The problem is that the industry built its UBI infrastructure on opacity: opaque scoring models, opaque third-party data flows, opaque disclosure about premium-increase risk, and opaque recourse when scores are wrong.

That opacity made telematics programs extraordinarily difficult to trust even before the enforcement actions began. When only 28% of policyholders knew about their insurer's driver monitoring program — and only 24% of enrolled users had read their privacy policy — the "informed consent" that anchors behavioral pricing's moral legitimacy is effectively fictional.

The business cost of this opacity is compounding. J.D. Power's 2025 study found that while interest in UBI is growing, the purchase rate fell from 22% in 2023 to 17% in 2025, even as average auto premiums surged past $2,300 annually and cost-conscious consumers actively shopped for savings. Insurers built the perfect product for a high-premium environment and then undermined consumer trust in it so thoroughly that adoption is declining.

What a Trustworthy Usage-Based Model Would Actually Require

The industry's Matteo Carbone and others project that telematics will become mandatory within five to ten years. If that trajectory holds, the current model is unsustainable — regulators will not permit mass mandatory enrollment into an opaque surcharge system.

A defensible UBI architecture would require several structural changes the industry has resisted. Algorithmic transparency — not full model disclosure, but meaningful explanation of which factors drove a score and why — is the minimum viable standard for a product that can increase someone's insurance cost. Consumer data correction rights, as Maryland's SB 351 proposes, address the accuracy problem that drives complaint volumes. Strict limits on third-party data sharing, with explicit consent requirements for any secondary commercial use, would close the Arity/LexisNexis data brokerage loophole that regulators are now attacking through litigation. And opt-out without penalty — meaning a consumer who withdraws from telematics enrollment cannot be rated adversely for the act of withdrawal — would make participation genuinely voluntary rather than coercively structured.

Insurers that build toward those standards will have a durable product. Those still running "save up to 40%" marketing against a backdrop of FTC consent orders and state AG lawsuits are operating on borrowed time.