Automotive software lifecycles were built for a world in which requirements, architecture, design, code, and tests were created largely through manual effort. In the age of GenAI, that assumption is changing rapidly. Today, AI can help generate code, tests, trace links, and even draft engineering artifacts directly from feature intent, software use cases, interface definitions, and platform constraints.
This presentation argues that the industry should move beyond a document-centric interpretation of ASPICE toward an AI-native, evidence-centric development model. The focus should shift from manually producing every intermediate artifact to establishing optimized trusted generation flows, human-in-the-loop review gates, automated verification, and measurable assurance evidence.
The session will explore how requirements-to-code-to-tests workflows can be reimagined for software-defined vehicles while preserving safety, traceability, and accountability. It will offer a practical perspective on how GenAI can transform automotive software engineering from artifact-heavy process compliance to faster, smarter, and more effective assurance-driven development.
