Thought Leadership
The automotive industry is facing structural change, and AI could be the decisive lever.
Electromobility, geopolitical uncertainty, new competitors from China, the shift to software-defined vehicles, and the pressure to scale digital services profitably are impacting organizations whose structures have historically been geared toward hardware excellence. At the same time, the industry is experiencing an unprecedented wave of technological innovation. But there is a gap between strategy papers and productive deployment: many pilot projects are technically convincing, but few reach regular operation.
This contradiction has less to do with a lack of ambition than with a structural problem. Companies want to reap the benefits of modern AI, but their technological and organizational foundations are not prepared to deploy AI on a large scale in a secure, reliable, and business-effective manner. The Kyndryl AI Readiness Report 2025 quantifies this discrepancy: While 86% of executives consider their AI strategy to be “best in class,” only 29% see their organization as sufficiently prepared to manage risks in an AI-driven future. The lack of maturity is even more evident in terms of return on investment: only 42% report that their AI initiatives to date have generated any business value at all.
The cloud provides the foundation
The innovative momentum of modern IT infrastructures is further accelerating this development. The past few years have been marked by technological leaps that would have been virtually inconceivable without cloud-based systems. Generative AI and large language models rely on elastic computing capacities and GPU scaling in hyperscale environments. At the same time, cloud services enable departments to test new ideas faster than ever before without being slowed down by complicated IT processes.
But this democratization of innovation brings new challenges. Many large automotive companies pursue multi-cloud strategies that distribute data, services, and applications across different platforms. What was intended to provide security and flexibility often creates new silos. Data resides in different cloud stacks, departments access different “as-a-service” building blocks, and cross-functional governance is rarely consistently anchored. Numerous projects get off to a quick start but fail to find their way into a stable, productive landscape. This is hardly surprising: target architectures were not considered, data rooms are not interoperable, services cannot be rolled out securely across the company, and teams are simply not set up for the later operational phase. Gartner estimates that 70% of all cloud-based innovation projects fail at this very point. The cloud has accelerated innovation, but it has also revealed how difficult it is for companies to combine speed with structural stability.
Agentic AI is not just an upgrade
This is where the real challenge of modern AI becomes apparent. Agentic AI – systems that negotiate decisions, pursue goals, and dynamically control processes – cannot be grafted onto existing structures. It requires integrated data rooms, robust cloud architectures, clear governance mechanisms, and an organization that is willing to rethink decisions. My colleague Dr. Frank Becker, Kyndryl Consulting Partner for Data & AI, puts it this way: “Agentic AI is not an upgrade of classic AI, but an organizational paradigm shift. Anyone who tries to operate it within old process logics will inevitably reach its limits.”
Yet the benefits that scalable AI can bring are enormous and concrete. In vehicle development, simulations can be evaluated automatically, variants optimized, and decisions prepared. Development times that used to take months are shrinking to weeks. Supply chains, which are increasingly influenced by geopolitical factors and market volatility, are gaining a new form of resilience through agent-based systems: risks are not only predicted, but actively mitigated. The added value of AI does not come from the efficiency of individual steps, but from the quality, speed, and consistency with which complex decisions are made.
Practice instead of prototypes
International examples show that companies that are already successfully using AI in regular operations have first modernized their digital infrastructure. The BMW Group is working with Kyndryl on a global data infrastructure that forms the basis for software-defined vehicles, simulation, and high-performance computing. Stellantis is integrating IT modernization and business transformation to improve data availability and process integration in the long term. Mitsubishi Motors Europe relies on robust, modernized IT systems that form a reliable platform for digital services. Companies entrust Kyndryl with their critical IT environments and in return receive stability and security as well as competitive advantages through sustainably implemented AI and cloud technologies.
The next step for the automotive industry is not to produce more prototypes. It is to embed AI in such a way that it has an impact on the entire organization. This means realigning processes, integrating data rooms, orchestrating hybrid IT landscapes, and empowering teams not only to develop autonomous systems, but also to manage them responsibly. Companies that start doing this today are laying the foundation for speed, quality, and resilience – qualities that will determine future competitiveness in a global, volatile industry.
The automotive industry has enormous strengths: engineering expertise, operational excellence, and decades of experience in building safety-critical systems. Now it is time to combine these strengths with the ability to use AI not as a demonstrator, but as a factor of production. It is precisely this step that will determine who leads and who follows in the coming years.
