Nearly seven in 10 automotive organisations are using or piloting agentic artificial intelligence (AI) in product development, but weak governance, process inefficiencies and data-related challenges continue to hinder wider adoption, according to a report by the Capgemini Research Institute.
The report, Rewiring automotive product development: How organizations can pull ahead in the AI race, found that 33% of automotive organisations already use agentic AI, while another 35% are piloting the technology. Adoption in non-core engineering activities is expected to increase from 26% currently to 68% over the next two to three years.
The potential productivity gains are significant, with 83% of executives believing agentic AI will allow engineers to focus more on core engineering work instead of administrative and coordination tasks. However, only 5% of organisations reported having fully mature AI governance frameworks, highlighting gaps in data quality, traceability and accountability.
Agentic AI widens performance gap among automakers
Based on a survey of 800 senior automotive executives across 12 countries in North America, Europe, and Asia-Pacific, the report identifies agentic AI as a key driver of the next phase of automotive product development.
While AI investments are widespread, a select group of organisations is already generating greater returns. These high-value realisers reported 17% faster development cycles, 19% higher engineering productivity and a 19% reduction in time-to-market.
Their advantage stems not simply from faster AI adoption but from stronger data foundations, clearer business cases, robust governance and more disciplined operating models, the report said.
"Over the past decade, the automotive industry has undergone a profound transformation with the shift to software-defined vehicles, establishing the digital foundations that are now enabling the next wave of innovation. Agentic AI is creating a new performance divide in the automotive industry," said Laurence Noël, global head of automotive at Capgemini.
Organisations generating the most value are redesigning workflows, strengthening collaboration and establishing clear accountability for AI-enabled decisions, she added.
US and Europe lead in AI strategy execution
AI maturity varies significantly across markets, with the US and Europe reporting the highest levels of strategy execution and cross-functional planning. Among original equipment manufacturers (OEMs), 53% in the US and 46% in Europe are implementing AI initiatives. However, only 18% of US respondents and 16% of European respondents reported having a comprehensive cross-functional AI strategy, vision and roadmap.
In China, four in 10 respondents said their organisations were in the implementation phase, while just 7% had achieved a comparable level of strategic maturity. Implementation maturity remains lower in other Asia-Pacific markets.
Confidence in agentic AI depends on its application. Executives reported the highest comfort levels with its use in research and concept activities, at 59%. This fell to 40% for development activities, where AI-generated outputs can directly affect engineering outcomes, product quality and vehicle safety. The findings point to gaps in evidence, traceability and accountability mechanisms needed to deploy the technology at scale. Building trust will be critical to moving from experimentation to enterprise-wide adoption.
Vehicle data fails to drive continuous improvements
The report also highlighted gaps in converting real-world vehicle data into product improvements. Although connected vehicles generate large volumes of operational data, 64% of organisations said insights from vehicles in use primarily inform future vehicle programmes rather than upgrades to vehicles already on the road.
Over-the-air updates remain concentrated on cybersecurity and navigation, with fewer organisations deploying updates that modify driving functions or assisted-driving capabilities.
Scaling AI requires organisational changes alongside technology investments. Six in 10 executives said delays in adoption were primarily driven by process inefficiencies rather than technological limitations. Meanwhile, 45% of organisations had yet to define an AI strategy and roadmap for product development.
Collaboration across the automotive value chain remains another challenge as software-defined vehicles become more complex and development involves OEMs, suppliers, and technology partners.
Only 30% of organisations reported co-development models based on shared outcome metrics, while just 10% had established shared accountability across the product lifecycle.