NEXT HORIZONS 2026, a new cross-industry technology event hosted by Vector, brings together industry leaders, technology experts, and software engineers to explore the future of intelligent software ecosystems. Across two days, six keynotes, four Tech Tracks, and sessions featuring more than 50 speakers, the event will address the challenges of scaling software-defined systems, deploying trustworthy AI, and accelerating engineering while maintaining safety, quality, and control.
Intellias will join NEXT HORIZONS 2026, bringing its extensive expertise in automotive software engineering, software-defined vehicles, embedded systems, and AI-enabled development. As a Vector Certified Embedded Engineering Partner, Intellias collaborates with Vector on automotive technology initiatives, including a practical migration from C to Rust within AUTOSAR Classic environments, demonstrating how modern software technologies can be integrated with existing automotive architectures.
Qualifying AI Toolchain: An Open Problem
Make AI Deployable and Trustworthy: How can you make AI deployable while ensuring accountability, traceability and regulatory oversight?
In his session, Adam Konopa will explore one of the key challenges in AI-assisted automotive software development: How to qualify an entire AI toolchain operating across the automotive V-model. As AI agents become involved in requirements engineering, architecture, coding, verification, and compliance, their outputs form an interconnected evidence chain where errors can spread, accumulate, and become increasingly difficult to trace. While existing standards address AI deployed inside vehicles, qualifying non-deterministic AI tools used to develop and validate safety-critical software remains an open problem.
Drawing on Intellias practical experience with AI-assisted validation toolchains and IntelliKit, Automotive Technology Platform, Adam will examine why human-in-the-loop oversight alone cannot provide scalable and auditable assurance. He will discuss the potential of statistical tool qualification and cloud-based virtualization, as well as explain why qualifying individual AI agents does not establish trust in the complete toolchain. The session invites an open discussion on building trustworthy, evidence-based AI engineering processes.
Short topic highlights:
AI across the automotive V-model: AI agents are becoming part of requirements engineering, architecture, coding, verification, and compliance, creating a chain of tools that must be qualified.
Compounding errors and verification risks: Mistakes made by generative AI agents can cascade across development stages, while errors missed by verification agents may remain undetected throughout the process.
The AI toolchain qualification gap: ISO/PAS 8800 addresses AI deployed inside vehicles, but qualifying interconnected, non-deterministic AI tools used in automotive software development remains an unresolved industry challenge.
Meet Intellias at NEXT HORIZONS 2026 to connect with our mobility experts and discuss the challenges of AI-assisted automotive software development, trustworthy AI toolchains, and the future of software-defined systems.
