Why AI Agents Could Soon Ride Along in Your Rivian and Remember What You Told Them Elsewhere
: Rivian CEO RJ Scaringe says the automaker plans to let third-party AI agents, such as Anthropic's, run in its vehicles, with context carried over from conversations on other devices. No partnerships, launch date or privacy details have been announced, and driving controls will stay off-limits to agents.
Your car may eventually pick up a conversation you started on your phone. Speaking at TechCrunch Disrupt, Rivian CEO RJ Scaringe said the company plans to let AI agents from outside companies such as Anthropic run on its vehicles, according to TechCrunch.
"That's where it becomes magical," he said onstage. Rivian has not announced any signed deals, and examples such as Anthropic, OpenAI and Meta were mentioned as possibilities, not confirmed integrations.
The idea is that an agent could carry context between devices. Scaringe described a driver asking for ventilated seats and the car drawing on a conversation they had with an assistant elsewhere, as one account of his remarks puts it. A few details set the limits of what has been said:
- Rivian already runs its own agentic framework that works "like a traffic light," sending simple requests to a local model in the car and harder ones to cloud models, and it has been able to reach apps such as Google Calendar since May
- Scaringe said braking, steering and other vehicle-level controls will never be handed to an agent
- He gave no launch date and noted that bringing frontier products into the car will take significant coordination with AI labs, with details promised at Rivian's AI & Autonomy Day in December
The open questions are as notable as the vision. Scaringe did not explain how context would be shared between devices and the car, which is the privacy-sensitive part: what a driver tells an assistant in one place could follow them into a vehicle that other people ride in.
Rivian would also not be first, since Tesla added Grok to its cars earlier this year. As another report notes, substantial integration work remains for both the automaker and the AI labs.
The most important detail for drivers is the boundary Rivian has already drawn: these agents would extend conversation and convenience features, not driving, which means the real decisions to watch in December are about data handling and consent, not safety-critical control.
