Frame this as the edge case regulators are starting to take seriously: the danger is no longer only what the car does, but what the interface asks the driver to do while the car does it. On June 8, 2026, the National Highway Traffic Safety Administration submitted to the Office of Management and Budget an information-collection request titled "Drive-Mode Design Best Practices," seeking approval for what the notice describes as "a one-time voluntary experiment which will examine how different drive-mode implementations affect driver attention and performance compared to standard interfaces." It is, on its face, a Paperwork Reduction Act notice — the least glamorous category of Federal Register document. But the risk factor is where the honesty lives, and this paperwork is honest about something the industry usually papers over: that the design of the human-machine interface is itself a safety-relevant variable.
The procedural posture is routine. Under the Paperwork Reduction Act of 1995, any federal collection of information from the public has to be cleared by OMB, and the agency must publish notice and invite comment. This notice records that NHTSA already ran a 60-day comment period — a notice was "published on February 11, 2026" — and that "NHTSA received two comments." Two comments is a quiet docket, the kind of low-attention regulatory event that almost no one reads. That quietness is part of why it is worth surfacing: the substance is more consequential than the traffic suggests.
What 'drive mode' actually means here
"Drive mode" is doing a lot of work in this notice, and it is worth being precise about it. Modern vehicles increasingly present the driver with selectable modes — comfort, sport, eco, off-road, and, in vehicles with driver assistance, modes that change how much of the driving task the car handles and how it communicates that to the human. Each mode reshuffles the interface: which information is shown, how warnings are presented, what the driver is expected to monitor, and how control hands back and forth. The experiment NHTSA proposes is to measure, empirically, how those different implementations "affect driver attention and performance compared to standard interfaces." In human-factors terms, the agency wants to know whether some drive-mode designs degrade a driver's situational awareness or reaction relative to a conventional baseline.
That is a more pointed question than it appears. The central unsolved problem of partial driving automation — the SAE Level 2 systems like adaptive-cruise-plus-lane-centering, hands-free highway pilots, and the modes that toggle them — is the human-attention problem. A system that does most of the driving most of the time lulls the driver out of the loop, precisely when the system most needs the driver to be ready to take over. The interface is the only lever a regulator has on that dynamic. It cannot legislate attention directly, but it can study and eventually shape how the drive-mode design either supports or undermines it. This ICR is the data-collection front end of exactly that effort.
Why a voluntary experiment, and why now
The word "voluntary" in the notice signals that this is research, not a rulemaking — NHTSA is gathering evidence, not yet writing a standard. That distinction matters for how to read it. The agency is in the part of the regulatory lifecycle where it builds the empirical record that a future standard or guidance would rest on. "Best practices" in the title is the giveaway: NHTSA is positioning to define what good drive-mode interface design looks like, and a controlled experiment comparing implementations against a standard baseline is how it would generate the defensible evidence to do so. The strip-the-marketing read is that the agency is treating the HMI as a regulated safety surface in waiting.
The timing fits a broader pattern. As driver-assistance features proliferate across the fleet and the modes that govern them multiply, the variance in how manufacturers design those interfaces has grown — and with it the risk that some designs handle the attention problem far worse than others. A regulator that wants to address that variance needs comparative human-factors data it does not currently have. The two comments the February notice drew suggest the proposal slipped past most of the industry's attention, which is itself worth noting: the design rules that emerge from research like this can eventually constrain how every automaker builds its driver-facing modes.
The honest read
It would be easy to overstate this. The notice is a request to run one voluntary experiment; it is not a finding, a standard, or even a guarantee that a standard will follow. No specific interface is criticized, no manufacturer is named, and the collection is explicitly one-time and voluntary. The fair characterization is narrow and concrete: NHTSA is asking OMB for permission to study, empirically, whether the way vehicles implement drive modes helps or hurts driver attention relative to a conventional interface.
But for anyone tracking the safety of assisted and automated driving, the significance is in what the question concedes. The agency is acknowledging, in a routine paperwork notice, that the interface design is a variable that can move driver attention and performance — and that it intends to measure it. The hardest part of partial automation has always been keeping the human engaged, and the only durable lever on that is the design of what the driver sees and is asked to do. This ICR, document number 2026-11456, is the unglamorous first step toward turning that lever with evidence rather than assertion. The comment record and the eventual experiment design are the things worth following from here.