Ford spent 2026 setting two records that shouldn’t exist in the same year. It became the most recalled automaker in America, with 64 recalls covering more than 13 million vehicles — more than double what Stellantis, its closest competitor, racked up. And it became, for the first time in sixteen years, the highest-ranked mainstream brand in J.D. Power’s Initial Quality Study, the industry’s most-watched measure of how many things go wrong with a new car in its first 90 days of ownership.
Those two facts don’t cancel each other out. They explain each other. And that explains why Ford just created a new executive role, Chief Auto Safety Officer, and filled it with a 30-year federal safety veteran named David Hines.
On the surface, this reads like a straightforward defensive hire. Ford has been bleeding money and reputation to recalls for years, so it went out and hired one of NHTSA’s own to help make the bleeding stop. That’s true, as far as it goes. But it skips past the more interesting question: why does a company that just posted its best quality score since 2010 still look, on paper, like the most dangerous car company in America?
What Actually Happened
Hines starts Monday, based in Washington, D.C., reporting to Steve Croley, Ford’s chief policy officer and general counsel. His job is to run product safety compliance, governance and strategy across a team spanning engineering, digital design, quality, manufacturing, software and purchasing — essentially every department that can create, or catch, a defect before it reaches a customer’s driveway.
His résumé is the interesting part. Hines spent more than 13 years at NHTSA, most recently as director of the agency’s Office of Crash Avoidance Standards, where he helped write the Federal Motor Vehicle Safety Standards and steered the agency’s research into automated driving and driver-assistance systems. Earlier in his career, he was an on-site design engineer for Ford and its old electronics offshoot Visteon, where he helped develop the Ford 500 and Freestyle — two vehicles Ford discontinued after barely two years on the market. The man Ford just hired to fix its safety reputation has watched this company fail before, from two very different vantage points.
Why This Job Looks Suspiciously Like His Old One
Here’s where the hire stops looking like a PR move and starts looking like a plan. In November 2024, NHTSA hit Ford with a $165 million civil penalty — the second-largest in the agency’s 54-year history, trailing only the Takata airbag case — after finding Ford took too long to recall more than 600,000 vehicles with defective rearview cameras, and gave regulators inaccurate data about it along the way.
The consent order that came with that fine isn’t just a check Ford had to write. It’s a punch list. Ford must build a safety data analytics infrastructure, an end-to-end documentation platform for internal and NHTSA-initiated investigations, a VIN-based component traceability system, and a new imaging test lab, all under an independent monitor for at least three years. Ford also has to re-audit every recall it has filed over the past three years to confirm each one was properly scoped.
Read that list again, then read Hines’ old title again: director of the Office of Crash Avoidance Standards. Ford didn’t just hire a safety executive. It hired someone who spent over a decade on the other side of the table, evaluating exactly the kind of infrastructure the consent order now forces Ford to build. That’s not a scandal — it’s the same logic that leads banks to hire ex-regulators to run compliance after a settlement. You hire the person who already knows what the auditor is going to ask for.
The Paradox Nobody’s Talking About
Here’s the wait, really? part. J.D. Power’s 2026 Initial Quality Study measures owner-reported problems in the first 90 days with a brand-new vehicle: infotainment glitches, road noise, seat comfort, cupholder placement. Ford’s score improved to 152 problems per 100 vehicles this year, best among mainstream brands, ahead of Nissan and Buick — a jump of 41 fewer problems per vehicle than a year earlier, momentum the company has been advertising heavily in its own factory towns.
NHTSA’s recall count measures something almost entirely different. It counts vehicles, across model years going back to 2018, flagged for a safety-related defect or noncompliance discovered at any point in a vehicle’s life, through any channel: warranty data, field reports, a lawsuit, an internal audit conducted years after the fact. A 2019 model with a wiring harness that starts failing in year six shows up in this year’s recall tally. It has nothing to do with how well this year’s F-150 was screwed together.
That’s how Ford can be simultaneously the most recalled automaker in the country and, according to the people who actually buy its new vehicles, one of the best-built. The recall count isn’t a report card on today’s factory floor. It’s largely a report card on years of accumulated field data catching up all at once — which is exactly what a consent order is designed to force, on purpose, in one uncomfortable pile.
Recalls Have a Software Problem Now, Too
There’s a second wrinkle inflating the raw number. The single largest recall Ford issued this year covered roughly 4.38 million trucks and SUVs for an Integrated Trailer Module software bug that could disable trailer brakes and lighting while towing. Ford fixed it with an over-the-air update. Most owners never saw a dealership. Its 741,000-vehicle park-by-wire rollaway recall earlier this year worked much the same way.
Under federal law, both still count as full recalls, tallied exactly the same as a campaign that requires a technician to physically replace a part. NHTSA’s public numbers don’t distinguish between “we pushed a patch while you slept” and “bring it in, we’re replacing a suspension component”. As more of a truck’s safety-critical functions move into software, more defects will get caught and fixed remotely — and every one will still add its full vehicle count to a headline number that gets compared, unfairly, against automakers still fixing everything with a wrench.
What This Actually Means for Owners
None of this makes Ford’s recall streak a non-story. The company has also issued recalls covering far more vehicles than are actually broken — its Bronco fire-risk campaign covered 565,691 trucks over a wiring issue it estimates affects roughly 1 in 100 of them — and campaigns where the fix wasn’t even ready when the recall notice went out. Both carry real costs for real owners: resale value, dealership hassle, and the plain anxiety of a recall letter for a truck that’s probably fine.
That’s precisely what the consent order’s VIN-level traceability requirement is supposed to fix. Ford has already run into the opposite failure mode — a recent recall covering 255,000 Focus sedans admitted the company couldn’t prove a 2018 repair had actually been completed in the first place. If Ford can eventually identify which specific vehicles carry a defective part instead of recalling an entire production run to be safe, owners get fewer blanket recalls and more targeted ones. That’s a real, practical benefit buried inside a punitive federal order — and exactly the kind of system a career NHTSA standards director would know how to build to the government’s satisfaction, because he spent 13 years deciding what satisfactory looks like.
The Bottom Line
Ford’s underlying problem was never really that it built bad cars, at least not by the measure that governs how quality gets scored today. Its problem was speed: how fast it found defects, and how fast, and how honestly, it told the government about them. That cost it $165 million. It’s why an independent monitor will be watching for three to four years. And it’s precisely the skill set Ford just went out and bought.
The gray-beard engineers Ford rehired after its AI-heavy quality push fell short are fixing the assembly line. David Hines’ job is to fix the paper trail. A recall total has never measured how good a car is. It measures how fast a company caught itself — and for years, Ford wasn’t catching itself fast enough.