In 2021, more than 13,000 people died in alcohol-related traffic accidents. Attempting to minimize the pain and loss of human life caused by drunk driving is laudable, but the latest way that lawmakers are “doing something” about it may have significant consequences for drivers and passengers alike.

proposal introduced by Representative Thomas Massie (R., Ky.) but voted down 229–201 would have blocked funding for a federal law that requires passenger cars sold in 2026 onward to have “advanced drunk- and impaired-driving prevention technology.” This law typified gross government overreach when it first passed — almost two years ago, when it was sneaked into the $1 trillion Infrastructure Investment and Jobs Act of 2021. That it’s still going ahead, despite the as-yet-unsolved practical problems with mandating new technology, as well as serious privacy and security concerns, indicates a long road ahead for a bill that should have been DOA.

The text of the law leaves much to the National Highway Traffic Safety Administration (NHTSA) to decide. All the technology must do is “passively and accurately detect” if the driver’s blood-alcohol concentration is above the legal limit or “passively monitor the performance of a driver.” If the driver is found to be impaired, the technology must “prevent or limit” the operation of the vehicle.

Developing technologies that “accurately identify whether [a] driver may be impaired” will be hard enough in practice. The law obviously intends to prevent drunk driving, but “impaired” driving could include intoxication from marijuana, illegal substances, or prescription medication, as well as drowsiness and distractions from eating or cellphone use.

What forms of impairment will the technology be required to detect? If it must detect more than alcohol intoxication, then camera-based systems are likely to beat out systems that detect blood-alcohol concentration. What degree of impairment will be enough to “prevent or limit” the car from driving? What rate of false positives and false negatives will be acceptable? How will the car find a safe place to stop until the driver is able to drive safely?

As yet, none of these questions has been answered. Automakers are experimenting with air alcohol odor sensors, sensors on the gear shift that shine “infrared light through the driver’s fingertip” to detect blood-alcohol concentration, and cameras trained on the driver’s head and eyes. In 2022, Volvo introduced camera systems in its EX90 model, which demonstrates that the market can and will provide safety features for those consumers who wish to have help behind the wheel.

But edge cases may strain the technology that’s been developed so far, and automakers (and bureaucrats) will have to decide how to address them. If the fledgling development of autonomous vehicles is any guide, human errors in coding are likely to lead to system failures. If the system fails, does the car become inoperable — stranding drivers and passengers — or does it allow drivers to drive, making it easy and desirable to simply disable the sensors?

No doubt the NHTSA is wrestling with these problems (though the market is likely to do a better job than an executive-branch agency could). But working through the practical issues still leaves privacy and data-security concerns unanswered.

Will the car store data permanently? Will it be locally stored or uploaded to a server? Who actually owns the data: the driver or the automaker? Who will be allowed to access the data, and how will that data be protected? It doesn’t strain the imagination to think that bad actors such as hackers (let alone insurance companies or law-enforcement agencies) would love access to a permanent and continuous camera feed of a driver’s face — and the technology to shut down the car.

The best time to defeat bad laws is before they are passed — not two years after the fact. Unless Congress rallies to pull back funding for the research, the likeliest chance of killing the so-called kill switch will be the flexibility of the NHTSA to punt its rulemaking.

The secretary of transportation is required to define within three years the kind of technologies that would meet the law’s mandate. That three-year deadline is rapidly approaching. It’s likely that the NHTSA will release draft rules in 2024 for public comment, which will give the public an opportunity to weigh in and shape what these regulations look like. However, the NHTSA can extend the time period for three years so long as it provides annual reports to Congress. Automakers will have up to three years after the standard is issued to comply.

It’s unlikely that the Biden administration is dragging its feet in developing the rule. But, short of Congress getting its act together to repeal the law, one can always hope that the problems inherent in bureaucracy will delay implementation of a law so fraught with privacy and security concerns.