Rivian’s vehicles will be able to run AI agents from other companies, such as Anthropic, in the future, founder and CEO RJ Scaringe said this week at TechCrunch Disrupt. That could eventually let people pick up an AI conversation in the car after starting it on another device, without agents getting control of braking or steering.
“That’s where it becomes magical,” Scaringe said on stage. He gave no timetable for the change, and Rivian offered no further details. The company directed attention instead to its December “AI & Autonomy Day” event.
The company’s recently introduced AI assistant already allows drivers and passengers to talk to their electric vehicles. Scaringe said at Disrupt that the next step would be to add products from frontier AI companies. Claude, OpenAI’s Dots and Meta’s Muse assistant are possible additions, but Rivian hasn’t confirmed any of them.
Rivian skipped Apple CarPlay and Android Auto and built its own vehicle software from scratch. It also made many of the apps itself, including navigation, while working to bring in outside services like Apple Music.
Rivian says its in-house agentic framework figures out what a person wants and routes the request to the right part of the vehicle’s software. That could mean turning down the temperature, switching on the seat warmers or summarizing the news. The framework can also pick between different models to handle each request.
“Think of it as like a traffic light. So if you ask the vehicle something, it decides what model it’s going to use. It can use a local model that’s running on the edge and the vehicle. If it’s a more complex question, it can go up to the cloud. And then even for that, we have different models that we use,” Scaringe said.
As of May, Rivian’s agent could already reach external apps such as Google Calendar. Scaringe thinks the implementation would look more like an extension of how people use agentic apps today, with what an assistant knows about a user following them into the vehicle.
To illustrate, he described someone asking for the ventilated seats to be switched on, then moving to dinner planning. “And then simultaneously say, ‘We were talking yesterday about this meal. I’m planning to have guests over on Friday. Can you help me think through what we’re going to make?’,” he said. The car could answer using context from a conversation the driver previously had with that assistant on another device.
Making that possible would require Rivian to link its vehicles with the user context collected by apps such as Muse. Scaringe cautioned that integrating outside AI products into a car would be difficult, requiring substantial technical work and cooperation with the AI labs.
“There’s a lot of plumbing behind the scenes we have to do to make sure that that works, and it’s not just within Rivian. That’s also a lot of coordination with the labs to make sure they enable that plumbing,” he said.
Safety also restricts what the agents can do. Scaringe said Rivian must exercise extreme care when putting software near safety-critical systems. “You wouldn’t want an agent to determine, ‘hey, it’s time to apply the brakes, or time to steer the vehicle.’ So anything that’s vehicle-level controls are not being handed to an agent,” he said.
Rivian wouldn’t be the first American automaker to put AI agents in its vehicle software. Earlier this year, Tesla added a feature that lets users talk to both SpaceX’s Grok chatbot and its Grok Bot assistant.
