Rivian has laid out its long-term vision for self-driving and in-vehicle intelligence, revealing a new autonomy framework, custom-built chips, and an in-house AI assistant at its first-ever Autonomy & AI Day.
At the center of the announcement is Rivian’s next-generation Autonomy platform, designed to support increasingly advanced driver assistance today and fully autonomous driving tomorrow. The company says this stack will power future updates across its R1 and upcoming R2 vehicles, marking a major shift toward vertically integrated software and hardware.
Universal Hands-Free Driving Comes First
Rivian confirmed that its new platform will enable Universal Hands-Free (UHF) assisted driving on second-generation R1 vehicles. The feature will be delivered via a paid subscription called Autonomy+, launching in early 2026. Pricing will be set at either a one-time fee of $2,500 or a monthly plan costing $49.99.
UHF will allow hands-free driving on compatible roads, placing Rivian alongside General Motors, Ford, and Tesla in offering advanced hands-free systems. However, Rivian emphasized that this is just the starting point.
“Our updated hardware platform will enable dramatic progress toward Level 4 autonomy,” said CEO RJ Scaringe, outlining the company’s ambition to eventually deliver vehicles capable of fully autonomous operation in defined conditions.
Large Driving Model and a Path to Level 4
Powering Rivian’s autonomy roadmap is its Large Driving Model (LDM), trained like large language models. The idea is to move beyond rigid, rules-based systems toward a learning-driven approach that improves continuously as vehicles encounter new scenarios.
Rivian did not provide a timeline for Level 4 deployment but made it clear that the architecture is being designed with that end goal in mind. Level 4 autonomy, currently seen in services like Waymo’s robotaxis, allows vehicles to operate without human input within specific geographic areas.
Custom Silicon and a Lidar Bet
In a move reminiscent of Tesla’s strategy, Rivian unveiled its first custom processor, the Rivian Autonomy Processor (RAP1). Built on a 5nm process, RAP1 integrates processing and memory into a single multichip module and will power the company’s third-generation autonomy computer, known as ACM3.
Unlike Tesla’s vision-only approach, Rivian is doubling down on a multimodal sensor strategy. Future R2 models will include lidar, which Rivian says improves spatial awareness, redundancy, and real-time detection. The Gen 3 autonomy hardware stack is currently in validation, with production expected to begin toward the end of 2026.
Rivian Unified Intelligence and the New Assistant
Beyond driving, Rivian also introduced Rivian Unified Intelligence (RUI), its in-house AI platform. RUI will underpin a voice-based Rivian Assistant, launching in early 2026 on both Gen 1 and Gen 2 R1 vehicles.
The assistant will handle everyday tasks, connect with third-party apps like Google Calendar, and help Rivian improve service diagnostics and predictive maintenance. Notably, the system is model-agnostic, allowing Rivian to work with multiple large language models rather than locking into a single provider.
Market Reaction
Despite the ambitious roadmap, Rivian shares fell more than 6% following the event, reflecting investor caution around costs, timelines, and execution. Still, the announcements signal a clear message: Rivian intends to control its technological destiny and compete head-on in the rapidly evolving autonomy race.
What this really means is Rivian isn’t just building electric vehicles anymore. It’s building an intelligence platform on wheels.
