Papers One through Six documented extraction across six domains where the person generating value wasn't recognized beyond the service that acquired them. Each domain operated on a version of the same premise: the platform was digital, the participation was behavioral, and the value flowed outward to systems the participant never entered. Paper Seven examines a different kind of mechanism: the vehicle is a participation mechanism in the series that moves through physical space and reports back.
Every major manufacturer now ships vehicles with embedded cellular connectivity, GPS, and sensor arrays that record location, speed, acceleration, braking, cornering, idling, seatbelt use, door activity, and cabin audio, and in some cases interior cameras for driver monitoring. This data is transmitted to manufacturer servers and, depending on configuration and program participation, to dealer networks, insurance telematics programs, and in documented cases to data brokers. The driver doesn't negotiate these terms at the point of purchase. They're embedded in the vehicle's operating architecture and disclosed, if at all, in agreements whose length and complexity place them outside any reasonable standard of informed consent.
The domain examined in this paper is structurally distinct from every prior paper in the series for one reason. In Papers One through Six, the value extracted from participation flowed outward to platforms and was used to generate revenue from third parties. In Paper Seven, the value extracted from participation can flow outward and return to the driver as a price. The system doesn't only take, it can use what it took to charge the person it took it from.
The connected vehicle assembles participation data across several distinct categories, each with a different commercial destination.
Location and movement records capture where the vehicle goes, when, how often, and by what route. This data can reveal home address, workplace, medical facility visits, religious attendance, and social relationships with greater precision than any platform behavioral inference, because the vehicle is physically present at the location rather than inferring it from a digital signal.
Driving behavior records capture speed, acceleration, braking force, cornering, following distance, and related operating patterns. This is the data that insurance companies use to assess and price individual risk. It's the most consequential category in this paper.
Engine and systems data captures fuel consumption, maintenance patterns, diagnostic codes, and mechanical behavior. Manufacturers and dealers use this data for service targeting, warranty management, and product development.
Cabin data captures voice commands, audio and entertainment choices, and connected phone activity. In vehicles with interior cameras, cabin data extends to occupant presence and identity.
Device linkage and external identifiers capture phones paired via Bluetooth, app integrations, and connected accounts. This is how driver identity persists across systems and how vehicle behavioral data is linked to broader commercial profiles held by data brokers and insurers. A vehicle that knows where it went and a data broker who knows who was driving it are two parts of the same commercial product. The device linkage layer is what connects them.
Driving behavior combined with precise location, route patterns, and phone linkage doesn't produce anonymous data in any commercially meaningful sense. The combination of those signals, accumulated across trips and days, resolves to a specific individual with a certainty that increases as the dataset grows: a single trip is ambiguous but a pattern over weeks is not. The system doesn't need to identify the driver at the moment of data capture, it needs only to hold the data long enough for the pattern to emerge.
None of these categories is disclosed to the driver as a transaction in which they have recognized standing. Each is classified as operational data generated by the vehicle's systems rather than as a productive contribution originating in the driver's behavior.
General Motors operates OnStar, a connected vehicle services platform with approximately 16 million active subscribers in North America. OnStar provides emergency response services, vehicle diagnostics, navigation assistance, and connected vehicle features to subscribers who pay a monthly or annual fee for the service.
A New York Times investigation published in March 2024 documented that GM had sold detailed driving behavior data from OnStar subscribers to LexisNexis Risk Solutions and Verisk without consent that drivers understood or could meaningfully act on. LexisNexis compiled driving behavior reports associated with individual drivers that insurance companies then used to adjust premiums. Drivers whose data had been sold discovered that their insurance rates had increased based on behavioral information they didn't know had been collected, transmitted, or sold. Some drivers reported being unable to obtain insurance from certain carriers as a result of their LexisNexis driving behavior file.
Though GM stated it would halt the practice following public reporting, the mechanism had operated for years before the investigation made it publicly visible. Class action litigation followed the investigation, with plaintiffs alleging that the data sale violated consumer protection laws and constituted an unauthorized disclosure of personal information.
The system produced these outcomes regardless of what GM intended when it designed its OnStar data practices. The company didn't design a mechanism for raising drivers' insurance premiums, but a data monetization program. The pricing consequence was a system outcome, not a platform decision directed at any individual driver.
The GM and OnStar case is the documented instance of a system architecture that operates across the connected vehicle industry: vehicle behavioral data flows from manufacturer telematics infrastructure through data aggregators and brokers to insurance scoring systems that use it to price coverage. The case isn't an exception to the system. It's a visible instance of how the system functions.
GM OnStar's subscription revenue and subscriber base figures, available from GM earnings materials, provide a partial platform-side baseline for the PDR calculation in this paper. Even so, they don't capture the value of the data licensing relationships that the investigation documented.
This case is the clearest illustration in the series of participation data flowing directly into a pricing system that operates against the person who generated it. It isn't speculation about what platforms might do with behavioral data. Instead, it's a documented commercial practice whose consequences were discovered by the drivers it affected when their insurance bills arrived.
Tesla vehicles transmit continuous telemetry to Tesla servers by default, including location, speed, battery state, charging patterns, and driving behavior. In vehicles equipped with Autopilot and Full Self-Driving systems, telemetry includes, where enabled, camera-based environmental data from the vehicle's surroundings. Tesla has disclosed that this data is used to train its autonomous driving systems. Every driving decision a Tesla owner makes, every road condition they navigate, every edge case their vehicle encounters, is an input to a commercial AI development program as an embedded condition of normal vehicle operation.
Tesla's fleet learning model is explicit about its dependence on continuous behavioral contribution from every vehicle in operation. The company has described its fleet as the largest source of real-world driving data in the autonomous vehicle industry. That data is the primary input to a technology whose commercial value Tesla has built its autonomous vehicle business upon. The drivers contributing to that system are not recognized as participants in its development. They not only paid for the vehicle, they're also providing the training data for a technology platform whose value is measured in the billions, under terms that classify their contribution as a product feature rather than a productive input from a recognized origin.
The Tesla model differs from the GM OnStar case in one important structural respect. The GM case documented data flowing to third parties who used it for purposes adverse to the driver. The Tesla model documents data flowing to Tesla itself for purposes that benefit Tesla's commercial development. The driver isn't priced by the data in the same direct way, and they also aren't recognized as the contribution to a system whose value they helped build. The extraction in this case is developmental rather than immediately consequential. Its full commercial significance will be documented in Paper Thirteen, where AI training data is the subject of the series' final calculation.
This section establishes why Paper Seven is structurally distinct from every prior paper in the series.
In Papers One through Six, the value extracted from participation flowed to platforms and generated revenue from advertisers, data buyers, content investors, and subscription fees. The person whose participation was extracted received nothing in exchange for the behavioral contribution the platform captured. That was the misclassification: value originating in a person, captured by a system, with the origin unrecognized and uncompensated.
Paper Seven documents a different mechanism. The driver whose OnStar data was sold to LexisNexis didn't simply fail to receive compensation for a contribution the platform captured, the effect of the sale was reflected in their insurance premium. The driver generated the data. The data was sold. The data was used to assess the driver's risk profile and the risk profile produced a price. The driver paid the price.
Prior papers documented a loop that ran in one direction: participation out, value to platform, nothing back. This paper documents a loop that closes on the driver: participation out, value to data broker, pricing signal in, cost to driver. The driver is both the source of the data and the subject of its commercial application. In this domain, the application doesn't only ignore their standing as the origin, it uses their data to determine what they owe.
The scale of telematics-based insurance pricing isn't marginal. By 2024, many major United States auto insurers had active telematics programs. The Insurance Information Institute and individual insurer disclosures document premium differentials between drivers in telematics programs and those outside them, with behavioral risk scoring affecting premiums in both directions. Drivers whose behavior scores poorly under telematics assessment face premium increases that can be substantial. The behavioral data producing those increases originated in the driver but the driver had no recognized standing in the assessment process that used it.
Most drivers are required by law to carry automobile insurance. The data that can raise their premium isn't optional participation in an advertising system, it's a behavioral record assembled without meaningful consent, sold to a third party without the driver's knowledge, and applied to a legally mandated purchase. The closed loop is complete.
Once driving behavior data exists in the insurance scoring infrastructure, the pricing consequence isn't tied to any future action by the driver. The driver doesn't need to be re-identified, re-enrolled, or re-contacted because membership in the system is sufficient. The harm is a property of system presence, not of a specific transactional event.
The Introduction to this series established four conditions whose simultaneous presence is required for participation to function as voluntary exchange: survivable refusal, recognized standing, transparency of terms, and independent jurisdiction. Paper Seven shows these conditions failing through a mechanism that combines the hardware activation failure of Paper Four with a pricing consequence that no prior paper has documented.
Survivable refusal fails at the point of vehicle purchase in the same way it failed at hardware purchase in Paper Four. A driver who buys a connected vehicle and declines to activate connected services retains a vehicle with reduced connected functionality. For the majority of American drivers, who have no viable transportation alternative to personal vehicle ownership, refusal of vehicle ownership isn't survivable in any practical sense. The participation is activated by the purchase and the ignition.
Recognized standing fails because the driver isn't acknowledged as the origin of the data the vehicle generates. The vehicle is the product and the data is classified as a feature of the product's operation. The driver who generates the behavioral record through their use of the product isn't recognized as a contributor to the commercial systems that record and sell it.
Transparency of terms fails because the data sale to insurers wasn't disclosed in a form the driver could evaluate or act upon. The terms governing OnStar data practices were available in documents whose length and structure satisfy legal disclosure requirements. They don't satisfy the transparency condition the Personal Data Royalty (PDR) framework requires: sufficient knowledge of what participation generates to make a meaningful choice about whether to engage.
Independent jurisdiction is the strongest condition in this paper. The GM case has produced active class action litigation, FTC inquiry, and state insurance commissioner investigations. Each addresses a different dimension of the same mechanism, and yet, none establishes the driver as a recognized origin of the data whose sale affected their insurance pricing. They address the legality of the disclosure and the fairness of the practice, but they don't address the foundational misclassification this series documents.
The PDR calculation for Paper Seven is structured in two layers, reflecting the two distinct ways the vehicle behavioral record generates value: outward to the platform and back to the driver as cost.
In the Origin Economics framework, Y = λ · f(H, K, T) expresses output as a function of human-origin participation, capital, and technology, multiplied by whether the legitimacy conditions of the exchange were satisfied. Lambda fails at ignition. The driver didn't establish legitimacy conditions before data collection began. The vehicle recorded its first data point before the driver left the driveway, under terms they hadn't read, for purposes they hadn't been told about, flowing to parties they hadn't consented to reach.
Platform Baseline (secondary): GM OnStar's subscription revenue and active subscriber figures provide a partial platform-side value baseline. OnStar generates revenue through subscription services and, as the New York Times investigation documented, through data licensing relationships with LexisNexis and Verisk. This baseline captures what the platform realizes directly from connected vehicle participation. It's incomplete, because data licensing revenue is not separately disclosed, and it's indirect, because it doesn't capture the full value of the behavioral record across the data broker and insurance ecosystem. It's not the defining measure of this paper.
Pricing Effect (primary): The value of the behavioral record in this domain isn't most clearly observed in platform revenue, it's observed in the price assigned back to the driver. Telematics impact is observable in insurance pricing differentials documented in public rate filings and insurer program disclosures. State insurance commission filings, insurer telematics program documentation, and industry analyses of behavioral pricing differentials establish a range of premium adjustments attributable to driving behavior scoring. These figures represent the most direct available measure of what vehicle behavioral data is worth in the market that uses it most consequentially, because they show the price the driver pays for the data they generated.
The ledger entry for Paper Seven reflects this structure. It includes both a platform-side value baseline, partial and secondary, and a cost exposure range representing the premium differential attributable to telematics-based pricing. This is the first ledger entry in the series where participation produces a net negative to the driver: value extracted outward, cost potentially returned inward.
The safety objection holds that telematics data improves road safety, provides drivers with behavioral feedback, and enables insurers to price risk more accurately, which benefits safe drivers through lower premiums. This objection correctly identifies that telematics programs can produce premium reductions for drivers whose behavior scores well. It doesn't address the misclassification argument, which isn't about whether the data benefits some drivers. It's about whether the driver is recognized as the origin of the data and has standing in the system that uses it. A driver who receives a premium reduction because their telematics data scored favorably has still contributed behavioral data to a commercial system without recognized standing, without compensation, and without any role in determining how their data was assessed or by whom. In this domain, the same data flow is cited as both benefit and mechanism of pricing. The data that allegedly makes driving safer is the same data that can raise the driver's insurance premium. The benefit and the harm aren't separate transactions. They're the same one.
The consent objection holds that terms of service disclosed data collection and that subscribers to connected vehicle programs agreed to the terms governing data use. This objection correctly identifies that legal disclosure occurred. It fails for the same reason it has failed across every paper in this series. Legal disclosure in documents of sufficient complexity to deter reading does not satisfy the transparency condition the framework requires. A driver who activated OnStar for emergency response services did not meaningfully consent to their driving behavior being compiled into a commercial report sold to insurance companies to price their coverage.
The value exchange objection holds that OnStar provides emergency services, navigation, and vehicle diagnostics in return for connectivity, and that this constitutes a fair exchange. The existence of value on both sides of a transaction doesn't establish that the exchange has been accurately recorded or that both parties are recognized as contributors. The service has value. The participation has value. The exchange is mispriced because the participation isn't recognized as a contribution deserving compensation. It's classified as the operational byproduct of service delivery, captured without recognition of its origin.
The connected vehicle data sector faces regulatory exposure across several frameworks, developing in sequence from the trigger event that made the mechanism publicly visible.
The New York Times investigation in March 2024 was the trigger. It documented the GM and OnStar data sale to LexisNexis and Verisk with sufficient specificity to establish the mechanism on the public record and to make regulatory inaction difficult to justify.
Class action litigation against GM and OnStar followed the investigation. Plaintiffs alleged that the data sale violated consumer protection laws and constituted an unauthorized disclosure of personal information. The litigation is proceeding in federal court and represents the most direct available legal challenge to the specific mechanism this paper documents.
The Federal Trade Commission opened an inquiry into connected vehicle data practices following the investigation. The FTC's existing enforcement authority under Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices, covers data collection and disclosure practices that do not meet consumer expectations established by platform representations. The inquiry addresses whether automaker data practices meet that standard.
State insurance commissioners in multiple states opened investigations into whether telematics-based premium adjustments complied with state insurance rating laws. Those laws generally require that pricing factors be disclosed to applicants and that they be actuarially justified. A pricing factor derived from data the driver did not know had been collected and sold raises compliance questions under both requirements.
The Senate Commerce Committee conducted inquiries in 2023 examining automaker data collection practices across the industry. Those inquiries established the legislative record showing that the regulatory framework governing connected vehicle data has not kept pace with the data collection capabilities manufacturers have deployed. No comprehensive federal legislation governing connected vehicle data practices has been enacted.
The PDR calculation in this paper rests on confirmed figures from GM OnStar and observable insurance pricing differentials. It does not include the following participation categories. Each involves behavioral data generated through vehicle operation or connected mobility that isn't attributable to individual drivers at the level the formula requires. Their exclusion is methodological and their behavioral transactions are real.
Ford, Stellantis, Toyota, Honda, and every other major manufacturer operating connected vehicle programs collect driving behavior, location, and systems data under terms structurally identical to those documented in the GM case. None disclose per-driver data revenue at the level the formula requires.
Vehicle financing platforms and dealership networks integrate vehicle usage and service data into lending decisions, resale valuation, and marketing systems. A driver whose vehicle usage data indicates high mileage, deferred maintenance, or behavioral patterns associated with credit risk may find that data affecting the terms of their next vehicle purchase or financing arrangement. The vehicle behavioral record connects to financial products the driver must purchase, extending the closed loop documented in this paper into the financing relationship.
Third-party OBD-II telematics devices sold to drivers for insurance discounts make the data trade nominally explicit: the driver installs the device voluntarily in exchange for a potential premium reduction. The nominal voluntariness of this arrangement does not establish the driver's standing in the broader market their data enters once it leaves the device. They negotiated a discount with one insurer. They contributed to a behavioral dataset whose commercial circulation extends beyond that relationship.
Ride-sharing platforms including Uber and Lyft assemble continuous location, route, and behavioral records from both drivers and passengers. The driver whose vehicle and labor produce the platform's service is generating a behavioral record the platform owns. The passenger whose movement through the city produces origin-destination data is generating a mobility profile the platform retains. Neither has recognized standing in the commercial systems that data enters.
Fleet telematics systems operated by employers monitor driver behavior continuously, connecting vehicle participation to employment relationships. A commercial driver whose telematics record is poor faces employment consequences in addition to insurance consequences. The behavioral data generated through their work is owned by their employer and used to evaluate their performance, price their coverage, and in some cases determine their continued employment.
Navigation and mapping applications including Google Maps and Waze assemble continuous location and movement records from personal devices operated during vehicle use, extending the participation domain beyond embedded vehicle systems into device-based mobility tracking. A driver who uses Google Maps for navigation is contributing location and movement data to Google's commercial infrastructure on top of whatever data their vehicle's embedded systems are transmitting. The two streams are separate collections feeding the same commercial ecosystem.
The floor isn't the number. It's the portion that has been disclosed.
Every prior paper in this series documented extraction whose consequences were commercial: revenue generated for platforms, behavioral profiles sold to advertisers, content decisions informed by viewing records, purchase environments shaped by behavioral history. The driver who generated the data that raised their insurance premium didn't receive a recommendation or a personalized experience. They received a bill.
The misclassification this series documents isn't only that participation value flows to platforms without recognition of its origin. In this domain, the data they generated was used to price them. The system doesn't only extract value from the person who produced it. It can return that value as a cost. Prior papers established the outward direction of the loop. This paper establishes that the loop can close.
The price was never zero. In the connected vehicle, it was paid twice: once at the point of data generation, and once at the insurance renewal.
Running total after Paper Seven. Add only the lines that apply to you.
If you drive a connected vehicle
Platform-side value baseline: partial, not fully disclosed. OnStar subscription and data licensing revenue provides a floor that doesn't capture the full value of the behavioral record across the data broker and insurance ecosystem.
Telematics pricing exposure: variable, documented in state insurance commission filings and insurer program disclosures. Premium adjustments attributable to driving behavior scoring represent a cost that originates in data the driver generated.
Net effect: unrecognized contribution outward, potential cost increase inward.
Prior papers added unrecognized contribution figures to this ledger. Paper Seven adds a different category: a participation record that can be used to price the person who produced it. The floor established in prior papers measures what platforms realize from participation. The vehicle entry measures what the driver may additionally pay because their participation was sold.Kashmir Hill and Joann Muller, Automakers Are Sharing Consumers' Driving Behavior With Insurance Companies, The New York Times, March 11, 2024.
General Motors Company, earnings materials and OnStar subscriber figures, fiscal years 2022 through 2024. investor.gm.com.
LexisNexis Risk Solutions, consumer telematics and driving behavior data products. lexisnexis.com/risk.
Verisk Analytics, driving behavior data and insurance analytics programs. verisk.com.
Federal Trade Commission, connected vehicle data inquiry, 2024. ftc.gov.
United States Senate Commerce Committee, inquiry into automaker data collection and privacy practices, 2023. commerce.senate.gov.
Insurance Information Institute, telematics and usage-based insurance program data, 2024. iii.org.
Tesla, Inc., Annual Report on Form 10-K, fiscal year ended December 31, 2024, filed with the United States Securities and Exchange Commission, January 2025. Referenced for fleet learning and telemetry disclosure.
Imanol Arrieta-Ibarra, Leonard Goff, Diego Jiménez-Hernández, Jaron Lanier, and E. Glen Weyl, Should We Treat Data as Labor? Moving beyond Free, AEA Papers and Proceedings 108, 2018, pp. 38–42.
Eric Posner and E. Glen Weyl, Radical Markets: Uprooting Capitalism and Democracy for a Just Society, Princeton University Press, 2018.
Shoshana Zuboff, The Age of Surveillance Capitalism, PublicAffairs, 2019.