Rethinking Cockpit Control in the Era of Touchscreens and AI

Recent developments across regulation, Original Equipment Manufacturer (OEM) design strategy, and safety research point to a growing shift away from touchscreen-dominated vehicle interiors. The European New Car Assessment Programme’s (Euro NCAP) updated 2026 safety protocols now penalize vehicles that lack physical controls for key functions, including turn signals, hazard lights, horn, windshield wipers, and eCall, reinforcing the return of physical buttons following sustained customer backlash and usability concerns tied to touch-heavy interiors.

Supporting this shift, recent research from organizations such as the University of Washington and Toyota Research Institute indicates measurably higher driver distraction levels when interacting with touchscreens, including increased lane deviation and reduced interaction accuracContry. Together, these developments signal a broader industry recalibration of how drivers interact with increasingly digital cockpit environments.

The shift away from touchscreen-centric design reflects a deeper issue in modern cockpit development. Consolidating controls into digital interfaces has increased interaction complexity, rather than simplifying it. While touchscreens reduce hardware cost and enable flexible design, they also introduce higher cognitive load, longer eyes-off-road time, and limited tactile feedback, making even routine interactions more demanding for drivers. In response, OEMs are reorganizing cockpit interaction around a clearer hierarchy, where physical controls are reserved for safety-critical functions, touchscreens shift toward secondary or passenger use, and additional modalities fill remaining interaction gaps.

Within this evolving structure, Agentic AI is emerging as a complementary layer that enables more natural, intent-driven control of vehicle functions. Unlike traditional voice assistants that rely on discrete commands, these systems interpret contextual requests across domains such as navigation, climate, and media, reducing the need to navigate multi-layered menus. This does not replace physical or touch interfaces, but instead augments them by handling non-critical, multi-step interactions that are inefficient through manual input. Low-latency and privacy requirements continue to shape deployment, reinforcing the role of hybrid and edge-based architectures for in-vehicle intelligence. As cockpit systems become more software-defined, this layered interaction model becomes necessary to balance usability, safety, and system complexity.

Automakers and technology partners must treat control as a system-level design priority, aligning interaction models with safety, usability, and execution realities rather than relying on interface-led differentiation alone.

For automakers:

• Re-establish physical controls as the foundation for safety-critical interaction. Core functions such as turn signals, hazard lights, windshield wipers, and emergency controls must remain accessible through tactile inputs to meet safety expectations and reduce driver burden, with secondary functions such as climate layered appropriately.
• Adopt a hierarchical interaction model across modalities. Physical controls should anchor primary actions, agentic voice systems should handle contextual and multi-step tasks, and touchscreens should be reserved for secondary or passenger-oriented interactions.
• Design Agentic AI around intent orchestration, not feature expansion. Systems should prioritize coordinating actions across navigation, infotainment, and vehicle controls rather than introducing isolated capabilities that increase interaction complexity.
• Align agentic system deployment with edge-first compute strategies. Low-latency, privacy-preserving execution is critical for real-time interaction, requiring tight integration with onboard Systems-on-Chip (SoCs) and hybrid architectures.
• Ensure consistency across interaction layers. Drivers should be able to move between physical, voice, and digital interfaces without relearning workflows, minimizing friction and distraction.

For technology suppliers and ecosystem partners:

• Prioritize integration and safety validation over standalone capability. Agentic platforms must operate within automotive-grade constraints, including functional safety, determinism, and mixed-criticality environments.
• Enable cross-domain orchestration through standardized frameworks. Middleware and orchestration layers should support coordination across the cockpit, Advanced Driver-Assistance Systems (ADAS), and cloud systems while maintaining OEM authority over execution.
• Support OEM ownership of the control layer. As platforms such as CarPlay expand toward orchestration roles, suppliers should enable vehicle-native intelligence and integration flexibility rather than displacing OEM control of the user experience.

Control, not display, will define the next phase of cockpit innovation, as automakers that successfully balance safety, usability, and intelligent orchestration will be best positioned to differentiate in increasingly software-defined vehicles.