Stellantis wants Wayve's AI to shrink a 24-month cycle for hands-off driving systems

Filosa made the saving conditional on volume. "So as long as penetration follows the path that we believe is going to happen massively ... automotive scale, which is one of the complimentary assets that we bring to Alex, for instance, will decrease costs, which is very important, obviously," he said ^[3]. He was talking about level 2++ systems, which allow hands-off driving while the driver still supervises ^[4].

The schedule claim is harder to check. Filosa spoke in Turin on October 9 ^[1], days after reaffirming guidance for 2026, the current year ^[13]. That puts a North American launch in 2028 ^[5] roughly 15 to 27 months away, depending on the month. That is about one of the 24-month cycles he said Stellantis runs now ^[2][15]. The first product will reach buyers on something close to the existing timetable ^[15].

Any compression has to appear in the vehicles that come after it. Kendall said the Stellantis portfolio lets Wayve tailor driving characteristics to different vehicles on the same underlying AI system ^[12]. "One of the goals ... is the diversity of vehicles and brands that Stellantis have," he said ^[7]. The Turin cars were a Fiat 500e and a Maserati Grecale, after a Jeep Grand Cherokee earlier in the year ^[6]. If a second nameplate follows the first within months, the shared system has shortened the cycle. If each one takes close to two more years, Stellantis has a new supplier on its old schedule.

Wayve also works with Nissan and Mercedes and has said its technology is built to deploy across brands and markets ^[8]. In Filosa's words, scale is an asset Stellantis brings "to Alex" ^[3]. So Stellantis volume lowers the unit cost of a system two competitors also buy. Whether that gain shows up in Stellantis's price per vehicle or in Wayve's margin depends on contract terms. The reports include neither the terms nor a cost target.

Stellantis shares were at an all-time low when Filosa reaffirmed a 2026 target of a mid-single-digit rise in net revenues and a low-single-digit adjusted operating margin, qz.com reported ^[13]. The FaSTLAne 2030 plan commits 60 billion euros over five years across Jeep, Ram, Peugeot and Fiat ^[14], or about 12 billion euros a year ^[16]. Carmakers that set out to build these systems in-house have moved to specialist suppliers as costs rose and returns came slower than expected ^[9]. On a low-single-digit margin, I think buying the driving model is the sensible use of capital. Stellantis keeps ownership of the platform the model goes into, STLA AutoDrive ^[5].

Penetration could follow Filosa's path and bring Stellantis a lower price per vehicle. It could lag, partly because European regulators are still harmonising approval rules for these systems ^[10] while Tesla presses them to approve Full Self-Driving ^[11]. Or the volume could arrive and the saving could go mostly to Wayve, spread across three automakers ^[8]. I'd expect the schedule gain to show up first. It depends on one software system moving across Stellantis's own brands, while the cost gain depends on adoption and on how Wayve prices its software. If the second Stellantis brand on the system takes close to 24 months after the 2028 launch, that view is wrong.