Rivian recently introduced an in-house AI assistant that allows drivers to interact with vehicle functions through voice commands, handling tasks such as adjusting temperature, activating seat warmers, and summarizing news. Speaking at TechCrunch Disrupt, Rivian founder and CEO RJ Scaringe revealed that the company eventually plans to allow its cars to run AI agents developed by external labs, such as Anthropic. Scaringe did not provide a specific timeline for the rollout, directing attention instead to Rivian's upcoming AI and Autonomy Day event in December. Rivian has built its own in-house agentic framework functioning like a traffic light, which decides whether to process a driver request using an on-device local model or a cloud-based model depending on complexity. As of May, the framework gained the ability to access external applications like Google Calendar. However, integrating third-party frontier AI products requires extensive coordination with labs and careful engineering, particularly because Rivian must avoid handing over vehicle-level safety controls, such as steering and braking, to AI agents. Instead, Scaringe envisions implementations acting as extensions of existing agentic applications, potentially connecting vehicle software to personal user context collected across other devices. Rivian joins other automakers exploring generative AI integration, following Tesla's earlier addition of Grok chatbots to its vehicles this year.
AI agents are coming to Rivian’s cars
Rivian founder and CEO RJ Scaringe announced that the electric vehicle maker plans to integrate external AI agents from companies like Anthropic into its vehicles. The automaker has developed its own in-house agentic framework to route driver requests between local edge models and cloud models. While vehicle-level safety controls will not be handed over to AI agents, Rivian aims to connect its cars with broader user context and third-party AI assistants.
Why it matters
This move highlights the automotive sector's growing interest in integrating frontier AI models directly into vehicle software ecosystems. It also underscores the technical and safety challenges companies face when safely deploying third-party agentic frameworks near critical vehicle systems.