It started as a drunk driving prevention measure and grew, as these things tend to, into something considerably larger.
Buried in Section 24220 of the 2021 Infrastructure Investment and Jobs Act — a 2,702-page bill most Americans associate with road repairs and broadband rollout — is a requirement that will affect every car bought new in the United States from late 2027 onward. The National Highway Traffic Safety Administration must mandate "advanced drunk and impaired driving prevention technology" in all new passenger vehicles. If the system determines you are impaired, it can prevent ignition startup or restrict your vehicle's speed.
The law was passed with bipartisan support under President Biden. It survived into the current administration when President Trump signed the Consolidated Appropriations Act on February 3 2026, preserving both the funding and the legal requirement. It is not a pilot programme or a proposal. It is an active federal statute.
There is one significant problem: the technology required to fulfil it does not yet reliably work.
What the systems are supposed to do
The NHTSA has confirmed that the final standard will require these systems to "passively monitor a driver's performance to accurately detect if the driver may be impaired." In practice, that means two main sensor types working together.
The first is a passive breath sensor. Unlike the breathalyser you blow into at a checkpoint, this system samples the air in the vehicle cabin continuously and calculates the driver's blood alcohol concentration without any action from the driver. The research programme behind this technology is called DADSS — Driver Alcohol Detection System for Safety — and it has been running since 2008 under a partnership between automakers and the federal government. Seventeen years of development. Still not ready for mass production.
The second is an infrared camera system monitoring eye movement, pupil size, head position, drowsiness patterns and steering behaviour. This technology is more mature. GM's Super Cruise and Ford's BlueCruise already use cameras aimed at the driver to detect inattention during assisted driving. What Section 24220 does is extend that monitoring to every new passenger vehicle sold in the United States, making it standard equipment rather than a feature available only on premium models.
When the combined system detects impairment — whether from alcohol, drugs, fatigue or a medical event — it can lock the ignition before the journey begins or restrict speed during it. The car, in other words, becomes a gatekeeper.
Like this? Get the app: iOS | Android
The case for it
Impaired driving kills approximately 10,000 people in the United States every year. That figure has remained stubbornly resistant to decades of enforcement campaigns, public awareness programmes and lower drink drive limits. The argument for a technological solution is straightforward: if the car will not move when the driver is impaired, the death toll falls regardless of whether the driver intended to make good choices.
The 2021 law's sponsors and traffic safety organisations including Mothers Against Drunk Driving have made precisely this argument. MADD supported the provision explicitly, and the technology's proponents point to the fact that no existing enforcement mechanism catches every impaired driver before they cause harm. A passive system that operates continuously does not rely on a checkpoint being in the right place at the right time.
Automakers have been broadly supportive rather than openly resistant. The technology, if it works, integrates naturally with the data and sensor infrastructure that modern vehicles already carry.
The case against it... and the questions nobody is answering
The first problem is accuracy. Every detection system produces false positives. A passive breath sensor that misreads medication, mouthwash or a passenger's alcohol consumption as driver impairment, and locks the ignition accordingly, leaves a person stranded with no recourse. A fatigue detection system that flags an early morning commuter as dangerously drowsy because of their natural eye movement pattern creates a different kind of injustice.
NHTSA's own assessment has acknowledged that no currently available technology meets the accuracy threshold required for mass deployment. This is why the agency missed its November 2024 rulemaking deadline, to the "deep disappointment" of traffic safety organisations who wrote to express it. The September 2027 enforcement date remains on the statute books, but automakers will receive two to three years to achieve full compliance once the final rule is published. That timeline is already slipping.
The second problem is data. The law as written contains no provision allowing drivers to opt out. Drivers cannot choose not to be monitored in a vehicle covered by the mandate. What the law does not specify is where the biometric data goes after collection.
This is not a theoretical concern. Modern vehicles already transmit location data, speed, braking patterns and other telemetry to manufacturers. Insurance companies already use this data in some markets to calculate premiums. Law enforcement already subpoenas vehicle data in accident investigations. A system that now adds continuous biometric monitoring — eye movement, breath chemistry, drowsiness patterns — creates a dataset of entirely new sensitivity.
Startup Fortune reported that privacy advocates and civil liberties organisations are raising three specific concerns: the data could be shared with insurance companies and used to raise premiums, it could be accessed by law enforcement without a warrant under existing vehicle data precedents, or it could be sold to data brokers under the same loose frameworks that govern other connected vehicle telemetry.
None of these outcomes are prevented by the current language of Section 24220.
There are also bills in Congress seeking to repeal the mandate. None have passed. The law survives. The deadline approaches. The technology lags behind both.
What is already in your car
The surveillance described in Section 24220 is not arriving from nowhere. Many drivers already sit inside it without knowing.
GM's Super Cruise uses an infrared camera aimed at the driver's face whenever the system is active. Ford's BlueCruise does the same. Volvo has built drowsiness detection into multiple models. Subaru's DriverFocus system uses facial recognition to identify registered drivers and alerts if it detects distraction. None of these are features you can turn off. If you bought a vehicle with these systems, they came as part of the package.
What changes under the federal mandate is scale and consequence. The cameras become universal and the system's response shifts from an alert to a potential denial of service. Your car does not just warn you. It decides.
Where this goes next
The honest position, as Motor1 summarised it, is that the technology is not ready, the deadline will probably slip further, and the window before full implementation gives legislators time to revisit the mandate's privacy provisions if they choose to. That is the optimistic reading.
The less optimistic reading is that the law is on the books, the rulemaking process is underway regardless of delays, and the data governance questions that should have been resolved before the mandate was written are being left to a later rulemaking process that may not arrive before the cameras do.
Drunk driving kills 10,000 Americans a year. That number is real and it is not acceptable. The question the 2021 law does not answer is what it costs in privacy, accuracy and civil liberties to deploy a universal monitoring system that is not yet reliable enough to work — and who is watching the data once it does.
Sources:
- Motor1 — Your Car May Soon Be Monitoring Everything You Do Behind The Wheel
- Family Handyman — Federal Surveillance Tech Coming to All New Cars By 2027
- Startup Fortune — Your 2027 car may watch your eyes and decide whether you are fit to drive
- State of Surveillance — Your 2027 Car Will Decide If You Can Drive
- Townhall — There's a Horrifying Federal Law Set to Require Active Surveillance Tech in All New Cars by 2027
- Infrastructure Investment and Jobs Act, Section 24220 — Congress.gov