The latest report by the Capgemini Research Institute, titled "Reshaping Automotive Product Development: Taking the Lead in the Corporate AI Race," stated that sector companies are accelerating the transition of Agentic AI applications from pilot stages to actual product development operations. The report showed that nearly seven out of ten companies currently adopt this technology, split between actual deployment at 33% and pilot phases at 35%, with expectations for the adoption rate in non-core engineering activities to reach 68% over the coming years.
These developments reflect tangible values, as 83% of executive officers believe that agentic AI is capable of enabling engineers to focus on core engineering tasks instead of being occupied with administrative and coordination tasks. In contrast, a small percentage not exceeding 5% of organizations indicated that they possess a fully developed AI governance model, highlighting ongoing challenges associated with data quality, traceability, and direct accountability.
Performance Gap and Regional Variance
The report was based on a global survey of 800 senior executives in the automotive sector across 12 countries. It clarified that the companies achieving the highest value successfully accelerated development cycles by 17%, increased engineering productivity by 19%, and reduced time-to-market by 19%. This superiority is not attributed to the technology alone, but rather to possessing more robust data foundations and more disciplined operating and governance models.
In terms of geographical distribution, markets in the United States and Europe record the highest implementation levels of AI strategies, with 53% of original equipment manufacturers in the US and 46% in Europe already implementing AI initiatives. In contrast, four out of ten participating companies in China indicate they are still in the implementation stage, while the percentage of companies that have achieved similar strategic progress there reached only 7%.
What Do the Trust and Data Barriers Represent?
With the expanding scope of agentic AI usage, trust levels vary fundamentally depending on the stages of application; comfort rises to 59% in early-stage research and concept formulation, while it drops to 40% in actual development phases that directly affect engineering outcomes and safety standards. The report also pointed to an additional challenge represented by the difficulty of converting the massive data of connected vehicles on the road into immediate product improvements, as 64% of companies utilize lessons learned to guide future programs rather than modifying current vehicles.