Cameras are everywhere.
They watch Americans enter stores, walk through airports and approach their neighbors’ front doors. Police officers wear them. Businesses mount them above cash registers. Millions of people voluntarily turn cameras on themselves every day, taking selfies and posting videos documenting everything from vacations to dinner.
Now cameras and increasingly sophisticated sensors are moving into another space — the driver’s seat.
And that raises a question that goes well beyond drunk driving:
How much should your car know about you?
A provision buried within the massive 2021 Infrastructure Investment and Jobs Act directed the National Highway Traffic Safety Administration (NHTSA) to develop a federal safety standard requiring new passenger vehicles to contain “advanced drunk and impaired driving prevention technology.”
The objective is straightforward: stop impaired motorists before they kill someone.
But the technology necessary to accomplish that goal could also fundamentally change the relationship between Americans and their automobiles.
What Congress Actually Required
Section 24220 of the 2021 infrastructure law defines the technology broadly.
A system could passively monitor a driver’s performance to determine whether the driver may be impaired and “prevent or limit motor vehicle operation” if impairment is detected.
Alternatively, technology could passively detect whether the driver’s blood-alcohol concentration has reached the legal limit and similarly prevent or restrict operation. A vehicle could also combine the two approaches.
Despite claims circulating online, however, the law does not specifically require a camera in every vehicle, nor is there currently a finalized federal mandate requiring cameras or so-called government “kill switches” in every 2027 model.
NHTSA is still trying to determine how the congressional mandate can actually be implemented.
And that has proven difficult.
What Happens When the Computer Gets It Wrong?
NHTSA’s February 2026 report to Congress makes clear that significant technological obstacles remain.
The agency is studying multiple approaches, including breath- and touch-based alcohol detection and systems that monitor driver behavior.
Camera-based driver-monitoring technology can examine characteristics such as eye glances and facial features. Other systems can analyze vehicle behavior, including variations in lane position.
But detecting possible impairment isn’t enough.
The technology has to be extraordinarily accurate if a computer is going to have the power to prevent a sober person from driving.
NHTSA’s research highlights the magnitude of that problem. Even an extremely small false-positive rate, spread across the enormous number of trips Americans make each year, could produce millions of incorrect determinations.
NHTSA previously defined a false positive as a situation in which the system determines someone has reached the impairment threshold when the person is actually not impaired or is below the legal limit.
The question practically asks itself:
How accurate should a computer have to be before it can take away your keys?
Your Car Already Knows More Than You Think
Privacy presents another concern.
Today’s connected automobiles can generate enormous quantities of information about their drivers.
Where did you go?
How fast were you driving?
Did you brake hard?
Were you driving late at night?
Those questions aren’t hypothetical.
In January, the Federal Trade Commission finalized an order settling allegations that General Motors and OnStar collected, used and sold precise geolocation and driving-behavior information from millions of vehicles without adequately notifying consumers or obtaining affirmative consent.
The FTC alleged that GM’s Smart Driver program collected information including speeding, hard braking and late-night driving.
For some consumers, precise geolocation information was collected as frequently as every three seconds.
The information didn’t necessarily stay between the driver and the car.
According to the FTC, driving information was sold to consumer reporting agencies, whose reports could then be used by insurance companies to deny insurance or determine rates. Some consumers complained after discovering that their driving habits were affecting what they paid for insurance.
The FTC’s final order now requires GM to obtain affirmative consent for certain connected-vehicle data collection and gives consumers additional options to access, delete or opt out of certain data collection. It also prohibits GM from sharing covered geolocation and driver-behavior data with consumer reporting agencies for five years.
That case changes the privacy debate considerably.
The question is no longer simply what automobile data might someday be used for.
We already have an example.
Maine Has Already Recognized the Problem
Maine law provides significant protections for information stored by traditional vehicle event data recorders.
Under Title 29-A, Section 1972, covered information generally cannot be retrieved by someone other than the vehicle’s owner without consent, although the law contains exceptions for circumstances including court orders, certain law-enforcement access, emergency response, vehicle repair and civil or criminal discovery.
But Maine enacted that protection two decades ago.
Automobiles have changed dramatically since then.
Maine has more recently addressed another aspect of connected-car technology.
State law requires the Attorney General to establish a telematics notice for prospective vehicle owners explaining telematics systems and describing the mechanical information those systems collect, store and transmit. Prospective owners must sign a certification saying they have read the notice.
But mechanical repair information is considerably different from technology potentially analyzing a driver’s face, eyes and behavior.
The question for Maine lawmakers is whether privacy protections written for yesterday’s automobiles are sufficient for tomorrow’s.
From Drunk Driving to Something Else?
There is another issue policymakers may eventually have to confront: mission creep.
Suppose a camera is installed to determine whether a driver appears impaired.
Could software someday also analyze distraction?
Fatigue?
Cellphone use?
Could increasingly sophisticated algorithms draw conclusions about a driver’s physical condition or behavior?
And, critically, what happens to that information?
The existence of a camera doesn’t necessarily mean video has to be permanently recorded. A properly designed system could potentially analyze a driver’s movements inside the vehicle without saving or transmitting the underlying images.
That distinction could determine whether the technology is viewed as a safety feature or surveillance device.
But motorists deserve answers.
Does anything leave the vehicle?
Is video stored?
Who owns the information?
Can manufacturers access it?
Can insurance companies?
Could police obtain it?
Could it be subpoenaed during litigation?
Does a camera see passengers?
Can the owner turn it off?
And what happens when a future software update changes what the system can detect?
Safety Versus Surveillance
There is an important argument on the other side.
Alcohol-impaired driving kills thousands of Americans every year.
Congress found when passing the 2021 law that alcohol-impaired driving accounted for roughly one-third of highway fatalities and cited estimates that advanced prevention technology could potentially prevent thousands of deaths annually.
If a vehicle can reliably determine that someone is intoxicated before that person pulls onto the Maine Turnpike, that technology could save innocent lives.
The policy question isn’t whether drunk driving is dangerous.
It plainly is.
The harder question is how much privacy Americans should surrender in exchange for preventing it.
We already carry GPS-equipped smartphones almost everywhere. We post photographs on Facebook and Instagram. Doorbell cameras capture people walking down public sidewalks. Dash cameras record highways.
But there is an important difference between choosing to post a selfie and purchasing a vehicle containing federally required technology capable of continuously evaluating the person behind the wheel.
NHTSA’s work is ongoing, and the agency has not finalized the standard Congress envisioned in 2021.
That gives policymakers an opportunity to establish the privacy rules before the technology becomes ubiquitous rather than after.
Because the ultimate question may not be whether a camera can determine if you’re safe to drive.
It may be something much simpler.
You bought the car. You pay the insurance. You pay the registration.
So who owns what your car learns about you?
big bother is watching you ……….