Stellantis Turns Yearlong Projects Into Weeks With AI — While Its Stock Sits Down 60%

By Nathan Brooks

Stellantis engineers built a hands-free driving system for a Jeep Grand Cherokee in six weeks — a project that used to take 12 to 18 months. That's the headline number from two days of interviews at the Chrysler Technology Center in Auburn Hills, Mich., where the automaker's engineers told Fortune that AI has turned yearlong projects into weeks and overnight simulations into minutes.

The company wants that speed to cut development time on its STLA Brain and STLA SmartCockpit software from four years to 24 months, starting with two models in 2027. But nearly every engineer who described a speed-up also described checking the machine's work. And Stellantis has bigger problems than time savings.

What does the speed actually change for practitioners?

Joydip Saha, global engineering manager and owner of Stellantis' L2+ hands-free highway assist, test-drives every new software build himself, sometimes twice a week. His team's six-week hands-free build came with AI tools. His rule on automation: "I will not let people use AI to architect stuff because they will forget how to debug stuff." In a safety-critical business, he says, engineers have to be able to answer every question.

Ben Zawodni, manager of software engineering, is folding hundreds of electronic control units into a central "brain" — STLA Brain. A proof of concept for a global observability platform took a week; before agentic coding tools, he remembers similar work taking six months. Defect fixes shrank from two to three weeks to two to three days.

Arturo Guzman's aerodynamics team runs design changes through an AI model that returns a drag estimate for a Ram pickup's front splitter in about three minutes. The full simulation used to take 14 to 24 hours. "There's no miracles. There's no magic," he tells vice presidents. "For us being engineers, all this is a Python code."

The financial backdrop is ugly

Stellantis posted a second-quarter adjusted operating margin of 1.8%, and CEO Antonio Filosa has set a 7% target for 2030. The company has 14 brands and a European business that is losing money. Europe accounted for about 38% of 2025 revenue — the most Europe-exposed of Detroit's Big Three.

In February, Stellantis wrote down roughly €22.2 billion (about $26.5 billion), more than Ford's $19.5 billion charge in December or the $7.6 billion GM took across two charges, after the $7,500 federal EV tax credit ended on Sept. 30, 2025. Shares fell 25% that day and are down nearly 60% year-to-date.

Mel Roberts, director of North American infotainment, said AI wasn't her biggest surprise. "What I couldn't anticipate is just the complete collapse of the electrification ambitions and how that really did affect all of our future plans," she said. "It trickled into everything."

Culture: redundancy, ownership, small teams

Chief Engineering & Technical Officer Ned Curic — formerly of Amazon, Microsoft and Toyota — said some teams show "phenomenal successes," but "as a whole organization, we are not there." His hardest sell: "the best thing that you can do for your career is to make yourself redundant."

Curic brought Amazon's two-pizza teams and single system owners to Stellantis. "If you see something that's wrong, you just pick it up," he said. "I own it." At the Paris auto show in October, he told Bloomberg the company would add "multiple hundreds" of engineers in Europe and the U.S. this year.

Partnerships over building everything

Anne Laliron, who runs a roughly 400-person tech research group, cuts about 10% of her group's projects each year. "If you want to embrace everything," she said, "you are at risk of spreading your effort." Stellantis designs battery cells in-house, partners with startups — Wayve on perception software, Tiamat on sodium-ion batteries, Zeta Energy on lithium-sulfur — and holds a stake in China's Leapmotor.

Its biggest bet: a battery-electric Dodge Charger prototype with a solid-state battery from Factorial, built with $250 million in capital. Factorial co-founder Siyu Huang puts the technology three to five years from automotive production.

Not everyone is speeding up

The wind tunnel still tests "typically, one car a day," said supervisor Steve Kubiak, a Chrysler veteran since 1992. Mohammed Malik's Windsor simulator can test 10 tires in an hour, but AI stays "in the box" — he uses it to find which data channel to inspect when a car "swirls a little bit." Roberts' AI research tool returns 10 to 15 sources she must verify: "I don't think it saves time," she said, though the work is more robust.

Alison Rahm, chief vehicle engineer for midsize and large SUVs, runs AI checks on roughly 1,000 engineering design practices. Trust takes time — her estimate for full confidence in AI is "four years," landing in 2030, the same year Filosa has promised that 7% margin.

Zawodni works at least twice as hard as before the tools. "Agentic tools, they've made our work so much easier," he said. "So why am I working so much harder?" The practical question for Stellantis is whether AI-driven speed can outrun a 1.8% margin, a 60% stock decline and a four-year trust curve — all converging on 2030.