Automotive Connectivity in 2026: The Conversation Has Moved Past the SIM

In conversation with Stéphane Lagresle – Host, Automotive Storytelling Podcast

The connected vehicle market is no longer a forecast. Infrastructure decisions that will define automotive connectivity for the next decade are being made in procurement rooms right now, and the gap between what enterprise buyers are asking for and what most connectivity providers can deliver has never been more apparent.

I recently had the opportnity to speak to Stéphane Lagresle, the host of the Automotive Storytelling podcast. With two decades of experience across Pioneer, TomTom, and Harman, he works closely with OEM engineers, Tier-1 product leads, chipmakers, and standards bodies on the questions shaping the industry’s direction. The central question I brought to the conversation: What does enterprise-grade connectivity for vehicles look like in 2026 and why are so many connectivity providers still struggling to win enterprise automotive business, even as the market grows?

Where do you think the automotive connectivity conversation has actually moved to?

Most of the connectivity conversation still happens at the SIM layer, Stéphane says. Roaming, coverage, eSIM migration, multi-IMSI. Those things matter. But the people he talks to, the OEM, the telematics platform, the fleet operator, have moved on to a different set of questions.

He maps three forces driving that shift.

The first is regulatory. The EU Data Act became applicable in September 2025, giving vehicle owners the right to access the data their cars generate and share it with third parties. From September 2026, a further obligation applies: connected products must be built so that access is possible by design.

“For an OEM, that is not a compliance footnote. It is an instruction to architect where data lives and who can reach it. If your traffic routes through a distant gateway before it reaches the application, that is a problem you have today, not one you can schedule for later.”

The second is the software-defined vehicle. Cars now receive updates over the air: ADAS refinements, security patches, feature changes.

“If the pipeline is unreliable, a critical security patch doesn’t reach the fleet, or an ADAS refinement sits undeployed across two hundred thousand vehicles. That is a brand liability at minimum, and a safety exposure at worst.”

The third, the one he thinks is most underrated, is operational visibility.

“For an OEM with a large portion of its latest vehicles connected, those cars form a fleet. Not in the traditional sense managed by leasing companies, but a fleet the manufacturer can access, collect data from, and push data to. When something goes wrong, the first question is what happened on the network. In most deployments today, the answer is that you wait for your operator to investigate. That is a significant operational liability when you have two hundred thousand vehicles on the road.”

You were at MWC Barcelona this year. What did that tell you about where the market is heading?

“Satellite connectivity for cars went from a slide in a deck to a product you could touch. A Tier-1 had a working satellite demo in an actual vehicle. You could press the SOS button and it placed the call over satellite. BMW showed their non-terrestrial network roadmap. NTN stopped being a buzzword this year.”

What interested him was not the technology itself but the conditions vehicle manufacturers attached to it.

“They were clear about two things. One, it has to be seamless, the car uses the terrestrial network until it hits a coverage gap, then switches to satellite with no new setup and no driver action. Two, it has to be standards-based, built into 5G through 3GPP Release 17, not a bolt-on with its own chipset. The moment you make the carmaker manage two separate networks, you have already lost the conversation.”

The conclusion he draws for connectivity providers is direct: this is not a SIM requirement, it is a network architecture requirement. The customer is asking for one continuous service across conditions that change by the second, and they do not want to see the seams.

He uses the Isle of Man TT races as an illustration. The most demanding road race in the world, villages, mountains, coast, riders going from two hundred kilometres an hour to almost nothing in seconds.

“The network of marshals around that race has to hold across every one of those conditions, with no degradation at the moment it matters most. A connected vehicle lives in exactly that world. Dense urban, rural motorway, tunnel, border crossing – all in one journey. The failures that count are the ones that happen at the wrong moment.”

What is the distinction most connectivity providers are getting wrong?

The most useful framing he heard this year came from a conversation about how to future-proof a car that will be on the road for a decade.

“The instinct, when a vehicle has a ten-year life, is to demand backward compatibility. Make sure whatever I buy today still works in eight years. But backward compatibility freezes your implementation. You are forced to keep carrying what you built first, forever.”

The better frame, he argues, is service continuity.

“You guarantee the service, not the implementation. The safety call has to work for the life of the vehicle. How it is delivered underneath can change across generations of radio technology, and the customer never has to notice. The mobile industry has done this before, voice has worked since 2G, through a completely different implementation each time.”

For connectivity providers, this reframe is the whole game.

“Your customer is afraid of lock-in, because a vehicle program runs six to ten years and you cannot re-architect it mid-life. The provider who wins is not the one who promises nothing will ever change. It is the one who makes change cost-effective and low-risk, so the service survives every generation underneath it.

You’ve talked about the supply chain changing. What does that mean for a connectivity provider in practice?

“For decades the model was a waterfall. The OEM wrote a spec and sent it down to a Tier-1, who dealt with the Tier-2s. Everyone knew their place. That model is over. What I heard all week at MWC was OEMs talking directly to chipmakers, to security companies, to software firms, building in the open, joining standards bodies, being explicit about what they need. The slicing of the cake is completely different now.”

The implication for a connectivity provider is uncomfortable.

“You are not going to get a clean spec handed down to you anymore. The capability that matters now is the ability to operate inside an ecosystem: to be in the conversation early, to be open, to be standards-aligned, to partner on equal terms. The providers who can do that will secure the enterprise automotive contracts.”

So what are buyers actually asking connectivity providers to deliver right now?

From the procurement conversations he is hearing, Stéphane identifies four requirements that come up consistently.

Traffic that can be isolated and reconfigured quickly. Isolation by vehicle program, by application, by region, changed in minutes rather than through a six-week change request. “Vehicle lifecycle management needs a flexibility the traditional APN model was never built for.”

Data that stays where regulation requires. Local breakout and the ability to guarantee that sensitive vehicle data does not leave the jurisdiction it belongs to. “After the Data Act this is not a nice-to-have. It is a gate you either pass or you do not.”

Visibility without a support ticket. The ability to see what happened on the network without waiting for the operator to investigate, and to act on it without raising a ticket. “For a global fleet, this is the difference between a problem you diagnose in minutes and one that costs you a truck roll.”

Processing close to the vehicle. Compute located near where the vehicle operates, rather than routed across the planet and back. “For ADAS, V2X, and anything autonomous, latency is not a performance metric. It is a safety parameter.”

“None of these are advanced features. In 2026 they are the baseline for any serious automotive deployment. The industry has to stop selling them as premium and start treating them as table stakes.”

How much time do providers have to position themselves?

“McKinsey predicts that more than ninety percent of new vehicles will be connected by 2030. The infrastructure question is being asked right now, by procurement teams at the biggest OEMs in the world. But the window to position is twelve to twenty-four months. The companies that close this gap now win the enterprise automotive contracts of the next three to five years.”

The demand Stéphane describes is the same demand we hear directly from operators and MVNOs building for the automotive market. The opportunity is real and it is live. These are active procurement conversations, not pipeline speculation.

What is equally clear is that the requirements have shifted decisively. This is not a coverage or roaming problem. It is an infrastructure problem: how data is routed, where it is processed, how quickly a network can be reconfigured around a vehicle programme, and how much visibility an operator can give an OEM when something goes wrong at two in the morning.

The providers who engage with that question on its own terms, and who can operate as genuine partners inside the automotive ecosystem rather than waiting for a specification to arrive, are the ones who will be in those rooms when the contracts are awarded.