Original research
Every vehicle software recall since 1994.
Since 1994 there have been 1,565 vehicle software recalls, affecting 115,460,436 vehicles. We have tracked every one of them.
- Years tracked
- 1994–2026
- Software recalls
- 1,565
- Vehicles affected
- 115,460,436
Why we built it
Because it did not exist.
Software is now the most common reason a vehicle is recalled, and the dataset that would let you see the trend clearly did not exist publicly. So we built it.
Every software-related vehicle recall since the first one in 1994, categorized by manufacturer, component, and vehicles affected — updated as new recalls are issued. We publish it because the argument we make to clients about software quality is more credible when the evidence is checkable.
Data compiled by Todd Warren, Founding Partner at Envorso, and published at autosofttoday.com.
Software-related recalls per year
Recalls issued, 1994–2026 · as of 2026-09-23 · 2026 year to date
Data compiled by Todd Warren, Founding Partner at Envorso, and published at autosofttoday.com.
Concentration
49 recalls carry 51% of the impact.
49 of the 1,565 vehicle software recalls on record — 3% of them — each reached at least 500,000 vehicles. Together they account for 51% of every vehicle ever recalled for a software defect.
- 49
- recalls affecting 500,000+ vehicles
- 59,370,054
- vehicles across those recalls
- 1,211,634
- average vehicles per recall
Largest single vehicle software recall in the dataset
Chrysler, 2018 — electrical system.
4,815,661 vehicles
NHTSA campaign 18V332000
Read the recall recordWhat it means
Software quality is a recall problem now.
A recall is the most expensive possible way to find a defect. It is also the most public. The trend in this dataset is not that vehicles became less reliable — it is that more of the vehicle became software, and software defects escape into the field in ways mechanical ones historically did not.
That is a delivery problem before it is an engineering one. The organizations that avoid it are not the ones with the best developers; they are the ones that find defects while they are still cheap, because their quality system, their toolchain, and their release discipline are built to.
That is the work we do.
Two hours to find out whether we recognise your problem.
No deck, no obligation. We listen, we tell you whether we have seen this before, and we say what we think it would take. If a Jump Start is the right next step we will say so — and if it is not, we will say that too.