International. Bosch and CARIAD, a software company of the Volkswagen Group, announced a new breakthrough in their Automated Driving Alliance with the joint development of an artificial intelligence (AI) software stack for Level 2 and Level 3 assisted and automated driving functions.
The solution, which is already being tested in vehicle fleets, will be ready for production projects from mid-2026.
The initiative seeks to make smarter and safer systems available to millions of drivers – from volume vehicles to high-end models, capable of imitating the natural behavior of a human driver. The software covers all key cognitive tasks: perception, interpretation, decision-making, and action.
The first versions of this technology are already running in test models such as the ID.Buzz and the Audi Q8. This year, hundreds of additional units equipped with state-of-the-art sensors will be added to the validation fleet operating in Europe, Japan and the United States. These vehicles collect high-quality data that allows the AI stack to be optimized and complex traffic situations to be addressed.
The Volkswagen Group plans to integrate these features into its new software-defined vehicle architecture, while Bosch will make the solution available to other manufacturers worldwide, thus promoting the global adoption of automated driving.
Peter Bosch, CEO of CARIAD, said: "We are demonstrating that the German automotive industry has mastered the key technologies of artificial intelligence and automated driving. With the expertise of our developers and engineers, we secure an integral part of Europe's digital sovereignty. Our goal at the Alliance is to make the convenience and safety of automated driving systems available to as many people as possible, so they can gain valuable time when they are in their car."
Mathias Pillin, CTO of Bosch Mobility, said: "When it comes to bringing automated driving systems to the roads reliably and at scale, data and AI are key. We can only overcome this challenge by working as equal partners and leaving behind entrenched mindsets. Together with CARIAD in the Automated Driving Alliance, we are showing how it can be achieved."
The alliance is applying AI to the entire technology chain: from object recognition and sensor fusion to decision-making and safe control of steering, brakes and powertrain. Inspired by generative AI models, the new stack is capable of analyzing complex urban scenarios and anticipating the behavior of other road users.
In addition, the architecture is designed to ensure safety, traceability and transparency in every AI action, with the future possibility of integrating multimodal vision-language-action (VLA) approaches, which mimic human reasoning and allow detecting hidden risks in traffic.
