Algorithmic Pricing, Surveillance Infrastructure, and Why Free Is the Only Honest Price
Manitoba just became the first province in Canada to ban personalized algorithmic pricing, the practice of using consumer data to charge different people different prices for the same product. Penalties include fines up to $300,000 for corporations and potential jail time for individuals. The bill makes clear that burying consent in a privacy policy doesn’t legitimize the practice.
This is being framed as a grocery store issue. It’s not. Or not primarily.
The sector most perfectly positioned for algorithmic pricing exploitation is EV charging. The infrastructure is already built for it. The data collection is already happening. The opacity is already normalized. And the consumer, often with a depleting battery and limited alternatives, is already captive.
Manitoba got there first with legislation. But one Canadian company got there first with a business model that makes surveillance pricing structurally impossible: charge nothing, know nothing, extract nothing.
Free turns out to be the ultimate transparency.
The Data Your Charger Collects
Every time you plug into a networked charging station, the system records: session start and end time, duration, energy dispensed, charging port ID, pricing policy applied, RFID card or account identifier, power draw patterns, current and voltage curves. That’s the technical data.
The app adds more: your payment method (which signals income band), your location history (where you charge, where you live, where you travel), your vehicle type (entry-level commuter or premium EV), your charging patterns (desperate fast-charger or organized home-charger), your price sensitivity (do you choose cheaper stations when available?).
Combined, this creates what privacy researchers call a user profile—and what Manitoba’s legislation calls the foundation for personalized algorithmic pricing. A network can infer not just who you are, but how much you’re willing to pay under specific circumstances.
The International Association of Privacy Professionals put it plainly: building profiles on individual users is one of the most aggressive uses of consumer information because it requires making inferences about individuals based on characteristics and past behavior. This can be perceived as creepy, invasive and in extreme cases discriminatory.
That’s the polite version.
How Algorithmic Pricing Would Work at a Charger
Imagine you’re driving through unfamiliar territory. Your battery is at 15%. The nearest charger is 10 kilometres away. There are no alternatives for another 80 kilometres. The charging network knows all of this: your battery state (transmitted by your vehicle or app), your location, the competitive landscape, your historical willingness to pay.
In a personalized pricing model, the algorithm sets your rate based on your desperation. Not peak hours versus off-peak. Not station-wide demand. Your demand. Your alternatives. Your inferred willingness to pay.
This isn’t hypothetical technology. Academic researchers are already developing personalized dynamic pricing policies for EVs using reinforcement learning, where each fast-charging station adjusts its charging price based on public and privacy-preserved information about the consumer. The question isn’t whether the capability exists. It does. The question is whether we allow it to be deployed.
What’s Already Happening
We don’t yet have a smoking gun showing exactly which charging networks are using personal data to charge different people different prices for the same session. As Vass Bednar, the Shield Institute managing director tracking this issue, notes: while firms may be eager to engage in this practice, they are less eager to be upfront with Canadians about it.
But the infrastructure and incentives are already in place:
Tesla’s Dynamic Pricing Pilot. Tesla is now testing real-time pricing based on live station utilization at ten North American sites. If the Supercharger is busy, pricing is high; otherwise, it’s low. Tesla frames this as efficiency. Users have been less charitable, one Reddit commenter said flatly, “This is a step too far,” while another wrote, “Surge pricing should be illegal.” Tesla’s congestion fees are already opaque enough that users report being charged premium rates at half-empty stations with no clear definition of what “busy” means.
ChargePoint’s AI Platform. ChargePoint’s new software platform explicitly enables AI-driven optimization that continuously analyzes usage patterns, energy supply conditions, charging station health and vehicle context to enable dynamic pricing strategies. Their system is designed to adjust pricing in real time based on individual session characteristics. ChargePoint doesn’t own the stations—they leave pricing to site hosts, which means thousands of independent operators can implement whatever pricing strategies they choose with minimal oversight.
Tiered Memberships as Price Discrimination. The entire membership model, EVgo’s three tiers, Electrify America’s Pass+, Tesla’s $12.99 non-Tesla membership—is price discrimination formalized. Charging management platforms explicitly advertise “dynamic tariffs by location, charger, and user group to unlock new revenue streams” and “different prices at the same location based on the type of customer.” This isn’t hidden. The data harvesting is the point: subscribers provide information on their charging habits, preferred locations, and peak usage times.
The Wild West Problem
Harvard researchers analyzing a million EV charging station reviews found that pricing in the sector resembles the “Wild West.” Unlike traditional gas stations, which often display fuel prices on lighted signs, EV stations rarely advertise what charging will cost. Drivers often arrive without any information on what to expect or how to make comparisons.
This opacity isn’t accidental. Research conducted by the same team found that charging station hosts, in the absence of regulation, have no incentive to share data, and they don’t. The lack of transparency creates perfect conditions for pricing practices that would be unacceptable if they were visible.
Consumer frustration is already high. One researcher noted: “Imagine if you go to a traditional gas station and two out of ten times the pumps are out of order. Consumers would revolt.” The EV charging sector has a 78% reliability rate, and pricing that’s even less predictable than functionality.
Manitoba Gets There First
Manitoba’s Bill 49 defines personalized algorithmic pricing as pricing based on the use of an algorithm or automated processing to set, recommend or vary a price offered to a specific consumer as a result of data about the consumer—including browsing or purchasing history, spending patterns, inferences about willingness to pay, geographic location, and demographic indicators.
The bill makes this practice an unfair business practice whether conducted online or in-store. Crucially, it specifies that consent buried in a privacy policy is irrelevant to the fairness question. You can’t sign away the right to non-discriminatory pricing by clicking “I Agree” on a terms of service document you didn’t read.
This is the first such legislation in Canada. A few US states have started moving on similar issues, but the policy space is wide open. Manitoba, as Bednar puts it, is helping kick-start a national conversation around something that firms would prefer we not talk about.
The timing matters. The US Federal Trade Commission has opened a probe into AI-driven personalized pricing. California’s Attorney General is issuing inquiry letters to retailers about their data practices and pricing experiments. Instacart quietly ended a program where customers saw different prices for the same product ordered at the same time from the same store.
The regulatory tide is turning. The question is whether the EV charging sector will be caught in it, or whether some operators will have built business models that never needed regulation to do the right thing.
The Free Alternative
There’s a charging model that makes algorithmic pricing structurally impossible: don’t charge consumers at all.
Plunk EV’s approach, free Level 2 charging, true plug-and-charge, no app required, no account creation, no payment processing eliminates the surveillance infrastructure entirely. If you’re not billing consumers, you don’t need their payment data. If you’re not optimizing extraction, you don’t need their location history or charging patterns. If there’s no price to personalize, there’s no personalized pricing.
This isn’t philanthropy. The business model monetizes environmental benefit through carbon credits rather than monetizing consumers directly. The revenue comes from the emissions displaced by EV adoption, not from maximizing what each individual driver can be convinced to pay.
The incentive structure is completely different. In a consumer-extraction model, the operator profits by charging more which creates pressure to identify when drivers are desperate and extract accordingly. In a carbon-credit model, the operator profits by enabling more charging, which creates pressure to make charging as frictionless and accessible as possible.
Free isn’t just a price point. It’s a structural commitment to not treating drivers as resources to be optimized.
The Frictionless Difference
I’ve written before about the three charging options near my local highway exit: the Plunk EV chargers at a family bakery with a petting zoo (free, no app), the DC fast chargers at a gas station development (premium price, app required), and the Level 2 charger at the community arena (pay-to-use, app required).
The Plunk chargers are almost always in use. The others sit empty.
Part of this is price. Free beats paid, obviously. But part of it is something deeper: the absence of surveillance. At the bakery, you plug in and go get a butter tart. No account to create. No payment method to register. No data trail documenting where you were, how long you stayed, what your battery state was, how desperate you might have been.
The charger doesn’t know you. It doesn’t need to. It just charges your car while you live your life.
This is what everywhere charging should feel like, incidental to daily activity, not an interruption of it. Invisible infrastructure, not surveillance infrastructure.
What’s at Stake
The EV transition is supposed to be about democratizing transportation energy. Anyone with access to an electrical outlet—which is nearly everyone—can “refuel” for a fraction of what gasoline costs. That’s the promise.
Algorithmic pricing threatens to recreate the inequities of the old system in new form. Instead of geographic price discrimination at gas stations (higher prices in low-income neighborhoods with fewer competitors), we get algorithmic price discrimination based on inferred willingness to pay. Instead of visible price gouging during emergencies, we get invisible surge pricing when the algorithm detects desperation. Instead of transparent market pricing, we get personalized extraction calibrated to each individual’s circumstances.
The communities that most need affordable EV charging, rural areas with few alternatives, lower-income neighborhoods with limited home charging options, remote regions where transportation costs consume disproportionate household budgets are precisely the communities most vulnerable to algorithmic exploitation.
Manitoba recognized this. The legislation isn’t about protecting consumers from a hypothetical future threat. It’s about drawing a line before predatory practices become normalized, before “that’s just how pricing works” becomes the accepted response to surveillance-based extraction.
The Path Forward
For policymakers, Manitoba’s approach offers a template: define personalized algorithmic pricing clearly, make it an unfair business practice, establish meaningful penalties, and specify that privacy policy consent doesn’t legitimize the practice. Other provinces should follow.
For site hosts considering EV charging infrastructure, the choice of network partner matters. Some platforms are building AI systems explicitly designed to optimize extraction from individual consumers. Others are building infrastructure that serves drivers without surveilling them. The regulatory trajectory is clear; choose accordingly.
For EV drivers, awareness is the first step. When a charging app asks for location permissions, payment data, vehicle information, and usage tracking, ask yourself: is this necessary to charge my car, or is this building a profile to extract maximum revenue from me? The apps that require the least are often the ones that respect you the most.
And for the EV charging sector as a whole, this is a moment of choice. The technology for algorithmic pricing exists. The data infrastructure is in place. The opacity is normalized. Nothing technically prevents the industry from following Uber’s surge-pricing playbook.
But nothing requires it either.
The model that’s actually working—at the bakery with the petting zoo, and increasingly at destinations across the country—doesn’t surveil, doesn’t discriminate, doesn’t optimize extraction. It just provides electricity to anyone who shows up, no questions asked.
Free turns out to be the most honest price of all.
* * *
This article is part of an ongoing series on EV charging infrastructure, energy equity, and the transition we should be building. Previous installments have explored the economics of Level 2 versus Level 3 charging, the vision of frictionless everywhere charging, and Canada’s Clean Fuel Regulations and industry compliance.
