The most accurate witness to how Americans actually drive isn’t a cop with a radar gun. It isn’t even a dashboard camera or the onboard computer in the car itself. It’s the phone in the cupholder.
That’s the idea of a September 2025 report from the Governors Highway Safety Association and Cambridge Mobile Telematics titled A Data-Driven Action Plan for Safer Roads, which is pushing state governments to retire the old way of deciding where roads are dangerous and replace it with a real-time feed of anonymized smartphone sensor data.
Currently, drivers with the highest levels of phone distraction are 240% more likely to crash than drivers who keep the phone in their pocket. Hard braking raises this by by 103% and speeding adds another 71%.
How the Data Gets Collected and What It Shows
Telematics sounds like surveillance jargon, but in practice it’s already running on most American drivers’ phones whether they know it or not. If you’ve opted into a safe-driver discount from State Farm, Progressive, Allstate, or Root, the accelerometer, GPS, and gyroscope in your phone are reporting back to an insurer in roughly real time. It’s picking up any phone handling, hard braking, sudden acceleration, speeding, and aggressive cornering.
Cambridge Mobile Telematics, the Massachusetts-based firm behind the GHSA report, processes this feed for most of those insurance programs through its AI platform DriveWell Fusion. It’s the largest telematics service in the world and claims to have helped prevent more than 100,000 crashes globally.
The old way of improving roads relied on crash data that was often two years out of date by the time planners saw it, and police reports chronically under-report distracted driving because drivers don’t volunteer that they were texting when they hit the guardrail. GHSA wants to change this using technology we all have in our pockets.
The report calls for less reactive and more proactive safety planning. for example, if they spot a section of road where phone handling is spiking, authorities can deploy enforcement or even road changes, then check the telematics data the next week to see whether anything actually changed.
Maryland is already testing this, with GHSA awarding the Maryland Highway Safety Office a $100,000 grant in late 2024, funded by General Motors, to use predictive telematics data to target distracted-driving hotspots across the state. Other states are following in their footsteps.
The Privacy Part That Actually Matters
Headlines you see on this sound pretty alarming.
Nobody at the Maryland Highway Safety Office is pulling up your name and issuing you a ticket for checking Instagram at a red light. The data that reaches state agencies is aggregated and anonymized. Planners will see that 500 devices confirmed phone-handling events on a specific stretch of Highway 9 between 7:30 and 8:15 a.m. on Tuesdays. They don’t see whose devices, and they certainly won’t receive phone numbers, names, or the contents of any messages.
The data currently comes from opt-in consumer programs, mostly insurance apps that offer discounts in exchange for letting the insurer score your driving. Cambridge Mobile Telematics strips the identifying information and sells the aggregated risk map to state highway agencies as a planning tool. What gets pulled isn’t your individual trip but the millions of trips happening on your road segment, which get averaged into a heat map.
Whether that’s dystopian or sensible depends on where you sit on the privacy-versus-safety scale, but the mechanics are worth understanding either way: aggregated and anonymized at the collection layer, contractually firewalled from law enforcement at the state layer, opt-in from day one.
Of course, this is a bit of a weird one. The devices that distract drivers are now the devices helping build and change the roads they drive on. But if it helps save lives, then we’re all for it.
