The Automotive Industry’s 4th Disruptor: Agentic AI in the Enterprise

“We are witnessing an explosion with agentic AI,” AWS CEO Matt Garman observed this week during his keynote at the company’s marquee annual conference, AWS re:Invent. “Every single customer experience, every single company and frankly, every single industry is in the process right now of being reinvented.”

Call me naïve, but coming into an AWS conference, I didn’t expect the entire two-hour keynote from their CEO, plus the second-day keynote from their VP of Agentic AI, Swami Sivasubramanian to focus on agentic AI. I thought I’d be hearing about traditional enterprise infrastructure and platform enablers like analytics applications, container management, serverless services, and so on, but no, just AI.

The strategy worked. You can now consider this fence-rider fully AI-indoctrinated. But I’m not sure that you could say the same for the rest of the automotive industry yet.

It’s an industry being disrupted thrice over by software-defined vehicles, electrification and autonomy, so you’ll have to forgive industry leaders, particularly those in organizations with distributed decision-making authority for struggling to embrace agentic AI. It might feel like adding agentic AI as the fourth axis of automotive reinvention could stretch R&D budgets to the absolute limit much less attention spans.

But the agentic AI reinvention has a distinct uniqueness to it that the others do not. SDVs, EVs, and AVs enable customer value. These are the table stakes that automakers must invest in and manage in order to protect their brands, their relationships with customers, and their positioning within the market. This calls for accelerated R&D efforts, capital expenditures, and restructuring efforts that reposition skills and supply chain partners to deliver on the promise of customer value.

AI, however, has a fundamentally different value proposition.

It’s Not A Chatbot, It’s a Minion 

“AI assistants are starting to give way to AI agents that can perform tasks and automate on your behalf… this is where you’re starting to see the returns that match up with the promise of AI,” Garman continued in his keynote address.

The first frontier of generative AI in automotive was the in-vehicle voice assistant, which made sense. When ChatGPT was released to the public in November of 2022, the potential for the technology to revolutionize conversational voice assistants became immediately clear. This value proposition extended to “copilot”-like applications that offer fast, natural-language access to information that helped people get things done faster. While not without its warts, the potential of the technology led to companies like Mercedes-Benz launching early versions of ChatGPT into their vehicles just one year later.

Agentic AI is different. Yes, these are still trained large language models, but through breakneck progress across the AI ecosystem from Nvidia’s hardware to Anthropic, Meta, OpenAI, Google, and others’ models and extreme capital investment in computing capacity, context windows expanded rapidly. This meant that AI companies could train their models with more and more data, leading to more accurate, robust, and specific outputs. By implementing agents that encapsulate specific context windows within enterprise environments, companies can now effectively replace monolithic applications with agents the end state for serverless computing architecture.

By creating agents that orchestrate workflows, managing inputs and outputs autonomously, CIOs all of a sudden see a future where, instead of developing and maintaining hordes of legacy applications, their developers can build agentic architectures that manage, improve, and triage in real time.

This is the new frontier of agentic AI, both for the automotive industry and for all industries not in the car, but in the cloud.

Now Is the Time for Automotive Leaders to Invest in Agentic AI

In front of a capacity crowd, leaders from Toyota Motor North America’s Enterprise AI and Toyota Connected organizations detailed their agentic AI journey, illustrating the rapid pace of innovation and optimization. Under the leadership of key automotive industry AI voices like Brian Kursar and Dave Tsai, Toyota has developed a template for how other automotive companies could successfully enable agentic AI across the enterprise.

Toyota’s ability to deliver on this vision has been many years in the making. By building and delivering on a strategy that prioritized in-house AI and engineering talent development, Toyota sits in a position of envy among many of its competitors—one that can lean on its skilled, diverse, and energized workforce to keep pace with rapid innovation from AI tooling leaders like AWS and its competitors.

This week, AWS launched a number of new agentic AI services aimed at enterprises with all levels of engineering capability, with some of the newest services - like Kiro - enabling spec-driven development. This lower barrier to entry means that automakers and suppliers facing increasing resource constraints and competitive pressure can quickly and easily lean into tools whose foremost value is enabling faster product development with higher quality.

This is the fundamental yet surprisingly quiet promise of agentic AI for the automotive industry. It’s not a new powertrain, supply chain, vehicle technology platform, or means of mobility. Rather, it’s a superpower that allows automakers to respond to disruptive pressures more efficiently and quickly, while enabling new pathways for innovation across the company.

Where Automotive Agentic AI Goes From Here

Multiple frontiers of innovation are emerging within the automotive industry as AI transforms what IT teams can do. Three that emerged from re:Invent this year include:

SDV Complexity — The biggest thorn in automaker SDV strategies is managing a complicated, fragmented mess of embedded software across architectures, regions, and even internal silos not to mention software suppliers. At re:Invent, General Motors demonstrated how it has been able to refactor legacy code using agentic AI, resulting in tangible improvements to cyclomatic complexity while maintaining adherence to standards like MISRA C.

Industrial AI — Building digital twins of factory floors and supply chains, using agents to monitor operations, predict failures or downtime, triage issues, and optimize or even support refactoring of business processes that enable more efficient and reliable factory operations.

Managing All That Vehicle Data — With automakers sensitive to any and all compute headroom built into cars, combined with the long operational lifespan of these assets, running agents in vehicles in the near term will be a challenge. But as OEMs deliver on SDV strategies and edge inference capabilities continue to improve, the possibility of agents running in the vehicle environment to support various customer- and OEM-focused data-intensive applications becomes increasingly realistic—improving customer privacy, lowering the cost of regulatory compliance, and enabling more predictive and personalized in-vehicle experiences.

Security AI — 2025 was a banner year - in a bad way - for automakers suffering from various cybersecurity challenges at the hands of black-hat hackers. Furthermore, vehicles are some of the hardest assets to secure due to their long lifespans, which often leave them with deprecated cryptographic capabilities just a few years into a program’s run. Agentic AI-enabled security agents in the cloud and in the vehicle could allow automakers to take a more proactive and scalable stance on cybersecurity. While industry leaders continue to invest in protecting their businesses and products today, the scalability of agentic AI could improve the resiliency of business operations many times over.

There’s no question that 2026 will bring more progress, and we’ll be keeping a close eye on how innovations in the AI space could affect automakers and their partners. In the meantime, we’re excited to continue working with our customers to learn from industry best practices, apply AI technology meaningfully and prudently, and identify the right partners to help them achieve true AI return on investment.

---

Next up? We’re back in Las Vegas for CES 2026. To set up time with our leadership team and analysts email us below:

Send us an email

To stay up-to-date on all the latest news about in-vehicle technologies, subscribe to our newsletter!

Subscribe Today