Automotive Outlook: 2026

The automotive industry stands at a crossroads entering 2026, facing a complex interplay of global tariffs, evolving electric vehicle (EV) dynamics, and the infusion of AI into just about everything.

As manufacturers and suppliers navigate recent financing shifts and regulatory changes, they also must address consumer concerns over EV affordability and range, OEM concerns over when to develop and implement new technologies, and security concerns that have plagued the industry for more than a decade.

“In 2025, the automotive industry was affected by tariffs, financial disruptions, and EV affordability challenges,” said Amit Kumar, director of product marketing and management for Automotive, Tensilica Product Group at Cadence. “AI-driven manufacturing, sustainability, and supply chain restructuring stand to define competitiveness in 2026. As the EV market was growing, we saw the $7,500 federal tax credit end in late 2025. This could potentially slow down EV market traction. Though I perceive this as a temporary blip, declining EV costs are a good sign amid growing competition between EV OEMs, providing more choice to customers and lower prices. Having said that, most EVs are still priced at approximately $40K, which is higher than conventional entry-level gas vehicles.”

Automobile financing costs dropped in 2025 due to lower interest rates, providing some relief for the auto industry. Now, OEMs are trying to balance profitability and consumer price sensitivity, while sifting through their supply chain base affected by recent disruptions. “2026 could be a balancing act between strategy and reality as OEMs balance tariffs, financing, and affordability,” Kumar noted.

Adiel Bahrouch, director of business development for silicon IP at Rambus, noted that the most substantive progress in 2025 was less about individual technologies and more about architectural and operating-model change. “Zonal E/E architectures gained real momentum as OEMs pushed to simplify wiring, reduce BOM (bill of material) complexity, and consolidate compute at the zonal and edge levels. In parallel, hardware-anchored security solutions also advanced as OEMs prioritized stronger protection for AI algorithms, data, vehicle computing platforms, in-vehicle networks, and OTA updates.”

Many advanced automotive technologies are still under discussion. They have yet to be adopted by OEMs and the tiered automotive suppliers. Even where new technology is showing up, the adoption rates can vary by region due primarily to varying economic factors.

“In the U.S., tariffs have essentially bought time for Detroit to try to institute something new technologically to catch up to, say, China,” said David Fritz, vice president of hybrid-physical and virtual systems, automotive and mil-aero at Siemens EDA. “China is really moving fast. China itself is struggling economically, which has slowed the introduction of other technologies. But they are being developed, and they exist. I’ve seen them, and what they can do is pretty mind-bending. Then, Japan is really trying hard not to get left in the dust — harder than any other region — and to keep up with what China is doing. We’ll see the results in the coming two years. Germany, meanwhile, is taking a very Swiss watch approach. It’s very fine-tuned, very highly detailed, very well planned, very well scheduled, not influenced by the speed of the rest of the industry. They’re doing proof of concepts, almost lab projects on the side, where they are looking into applying AI in some incredibly innovative ways to solve tiny pieces of the problem that, once assembled, one day will be a wonderful working Level 5 Swiss watch. That methodological approach is great, and we’re involved with what all of those companies are doing, and we are participating in those projects. In working with Germany, we’re learning an awful lot about how to democratize this technological move toward high-level autonomy. We’re also helping companies in Japan, for example, that don’t know how to get from Point A to Point B, to figure out how to do that using some of our lessons integrated into our PAVE 360 product.”

This year may be a bit of a wildcard, Fritz said, because many in the industry are still struggling to figure out when to develop new technologies, when to bring them to market so they don’t flop, and when they will result in increased sales, which is essential to justify the investments. “We can’t separate technology from the business side of things. They are intertwined.”

That has a snowball effect on other technologies, as well. “V2X technologies faltered as anticipated government mandates and funding were delayed, while attention shifted instead toward V2N-based (vehicle-to-network) services,” said Ken Horne, automotive and energy strategic planning at Keysight EDA. “In 2026, software-defined vehicle architectures are expected to gain momentum among traditional OEMs, enabled by higher-performance in-vehicle computing, deeper integration of AI and machine learning into vehicle decision-making, and the likely introduction of mandated radar testing within periodic technical inspections.”

Autonomy everywhere
This year also is likely to witness increasing overlap between automotive and robotics.

“Jensen Huang is talking about physical AI. Elon Musk has mentioned physical AI. And Arm recently announced they’re restructuring for physical AI, putting automotive and humanoid robotics in the same business unit,” Fritz said. “That makes absolute sense because we have been using robotics technology for the last five or six years on PAVE 360 to help automotive move in that direction.”

Cars and robots are on similar paths in the race toward full autonomy. This includes everything from wider deployment of ADAS to in-vehicle networking such as optical Automotive Ethernet, 10BASE-T1S, and new camera/display links. It also relies heavily on software-defined vehicle technology.

Cadence’s Kumar points to a “fundamental shift in the automotive industry and blurring the lines between traditional automotive engineering and advanced software development. SDV growth also benefits OEMs as it enables them to continuously monitor and enhance vehicle performance and customer experience long after the car has left the factory floor and creates monetization models for car makers.”

Unlike traditional vehicles, where functions are hard-coded into individual ECUs, SDVs utilize powerful central processors and interconnected software systems that allow over-the-air updates. The result is more features, functions, new applications, safety, and security, along with real-time diagnostics and seamless integration with cloud services.

At the same time, 2025 exposed the limits of software-only or compute-only SDV strategies. According to Rambus’ Barouch, programs struggled where the following were missing:

• Tight hardware and software interoperability;
• Hardware platform and software stack scalability, and
• Fast update and upgrade cadence.

“As a result, platform competition became tangible, not just around software features, but around hardware-anchored security, hardware/software integration, engineering velocity, cost structure, and long-term operational control,” he said, adding that the auto industry also managed multiple business pressures in 2025 by shifting from aggressive expansion to discipline and prioritization. “OEMs strengthened second-sourcing strategies, deepened partnerships with semiconductor suppliers, and reintroduced longer-term planning to handle residual supply-chain volatility. After several years of heavy investment in SDV platforms, autonomous driving, and electrification, many OEMs reached a financial inflection point. As a result, product roadmaps were shifted to better align with realistic adoption curves for electrification and autonomy, and software strategies became explicitly ROI-driven with a stronger focus on monetization and lifecycle value. 2025 also marked a broader industry recognition. Speed, largely driven by China, began to outweigh perfection in SDV markets. Faster iteration cycles and tighter hardware-software integration set new competitive benchmarks globally.

This year, Bahrouch expects the SDV strategy to be increasingly replaced by P&L accountability as OEMs demand clearer returns on software investments, relying more heavily on ecosystem partners for cost sharing and risk management, while intensifying collaboration with silicon suppliers to co-design scalable, cost-optimized hardware compute platforms. “The strategic focus will shift toward minimizing total system cost, while continuing to invest in safer, secure, and interoperable software stacks with update capabilities. Companies that balance disciplined procurement and engineering velocity with innovation in silicon and system design will be those that thrive.”

Navigating industry transformation
To understand where the industry stands today, it’s helpful to look at how other industries have navigated similar periods of transformation, especially given the complexities of the automotive ecosystem, both business- and technology-wise. Drawing parallels can offer insights into what lies ahead for automakers as they grapple with both technical and cultural shifts.

“Remember back when everybody had a Nokia flip phone? You just flipped the thing open, and it worked,” Fritz said. “Nokia had the entire industry in the palm of its hand, literally and physically, but why was Nokia so successful? Nokia’s advantage was its value chain, the number of partners it had. There were approximately 350 components in that tiny little circuit board, and the advantage was having that supply chain all set up and all figured out. It would take forever for other companies. Motorola and others tried, and they couldn’t put together the supply chain, so the value chain started breaking down, and Nokia owned the world. Then, a little bitty company in San Diego that starts with a Q came out with a system-on-a-chip. They could not sell that. They couldn’t pay to give it away. Why? People said, ‘What can I do with this one chip when I have 350 discrete components I have to worry about?’ Then everybody realized that with that one SoC they could replace 325 of those 350 components. ‘That sounds great,’ they said, ‘but how do I build a phone with that?’ What Qualcomm did was create something called a reference design so they would build smartphones, not for sale on the market, but to say, ‘I’m building a smartphone for you with our SoC. You take it, you take it apart, you play with it, you learn. And then you’ll realize you can build a smartphone with this and do a lot more than you could before.’ Then the customer would take that reference design, change it, and go to market with that. That was the downfall of the Nokia flip phone.”

This is where the automotive industry is today. “‘How do I go from 60 to 100 ECUs, built by 25 or 30 different companies with software from all these different teams, down into something that I can control, that I can specify, and can handle these AI workloads, in which my suppliers have absolutely no idea how that actually works? We talked about this in the industry years ago,” said Fritz. “Consolidation was happening, but it’s only now becoming a reality. And now, the OEMs and tiered suppliers finally understand. Software-defined vehicles in that context means, ‘I have this SoC, it does the work of five ECUs, and I can write the software. But that means that I need 64 or 128 CPUs, and I need an onboard GPU and a network on chip.’ They’re going through the transition that cell phones did quite a few years back, and landed on the solution of, ‘I need a full system-on-a-chip design that meets my needs,’ which implies you understand what your needs are. How do you figure out what your needs are? That’s where the cultural change and the big emphasis on new software teams and agile methodologies in automotive come together. In 2025, the world was just starting to understand that. In 2026, it will begin to solidify.”

Whether it’s wrapped around physical AI or software-defined doesn’t matter, because the fundamental concept is finally starting to resonate. “They can back it up with having the skill sets necessary to take advantage of that,” Fritz said. 

Automotive cybersecurity
As the industry consolidates around new hardware-software paradigms, attention also is turning toward another critical area — cybersecurity. With the growing complexity and connectivity of modern vehicles, safeguarding automotive systems has become an essential component of future innovation and operational stability.

“2026 is going to be a progression of some of the adoption that we saw in 2025,” said Dana Neustadter, senior director of product management for Security IP Solutions at Synopsys. “We’ve seen cybersecurity in the car with ISO 21434, which is now more or less a given. We’ve also seen interface security taking off, like for PCIe and MACsec. PCIe and Automotive Ethernet work together to provide that high-speed backbone for software-defined vehicles as a means of handling data requirements. There is a lot of data in ADAS systems, like in autonomous driving. Think about security and safety. So when we talk about an automotive hardware security module, when I talk about interface security, we talk about that together with functional safety.”

Post-quantum cryptography support also ramped in 2025, and will accelerate this year, Neustadter said. “Security in the vehicle will expand, because a lot of the data that moves across interconnects is sensitive. That requires more PCIe and Ethernet security, because PCIe helps with the low latency connections in-vehicle when you compute information, or when you have a complex sensor type architecture and Ethernet.”

MIPI security interfaces are expected to show up in more designs, as well. “We already see some lead customers adopting MIPI in vehicles for its camera serial interface (CSI). This is related to sensors and images that are being processed and sent to application processors in the car. More recently, the specification has been updated with security extensions, camera service extensions (MIPI CSE), which bring in security for authenticating, as well as encryption in some cases. So it provides confidentiality of the sensor data and uses a standard type of cryptography, like AES or  GCM. We expect this type of security to be adopted more in 2026.”

Further, based on customer adoption, Neustadter predicts an increase in the adoption of inline memory encryption for LPDDR-type interfaces in the car. “Again, it’s all about data being sensitive or private or important for the security of the car. LPDDR5 with inline memory encryption, LPDDR6 with inline memory encryption in the car. And here again, this is automotive grade, which means it is ASIL-B compliant. So functional safety is, at a minimum, ASIL-B, functional safety ISO 26262, but also the mandatory cybersecurity ISO 21434.”

Conclusion
The automotive industry is witnessing a broad transformation in vehicle computing and system architecture. As these innovations continue to gain momentum, their impact will reshape how vehicles process, secure, and utilize data for both safety and user experience.

“The rise of high-performance in-vehicle computing is especially exciting as AI moves deeper into the vehicle, driving compute-density, memory bandwidth, security, and data-movement requirements,” Rambus’ Bahrouch said. “Zonal control architectures are another compelling area since they require new approaches to data fabric design and fault-tolerant hardware. As AI-driven workloads are increasingly consolidated alongside safety and non-safety critical functions on shared hardware platforms, the industry is being forced to rethink safety, isolation, and security at a much deeper architectural level. This evolution is also accelerating the need for hardware-anchored security and robust update mechanisms to protect AI models, data pipelines, and vehicle behavior over the full lifecycle. Together, these trends create a significant opportunity for innovation across compute, memory, interconnect, safety, and security, rewarding teams that can solve system-level challenges rather than optimizing individual components in isolation.”

The upshot is that 2026 will see increasing traction for scalable hardware platforms targeting ADAS and AI workloads, and greater adoption of chiplet-based compute platforms, because they offer a practical path to scalable platforms and cost control. “Equally important, these platforms will increasingly be paired with interoperable, scalable software stacks with robust update capabilities,” Bahrouch said. “For central gateways, we see the move from ISO 26262 ASIL-B level support toward ASIL-D. High-bandwidth memory architectures should also see increased interest as vehicles demand faster data movement for sensing and AI workloads. For security, we anticipate accelerated adoption of hardware-anchored quantum-safe cryptographic algorithm support. We should also see traction in next-generation power management ICs that help automakers cope with higher voltage systems. Together, these trends will give hardware teams new ways to optimize cost, performance, and reliability at the system level.”

Related Reading
2025 – A Year Of Change And Anticipation
Unexpected change rewards those who are quick to adjust, but all aspects of a system must be adjusted to keep growth going.