Rivian shares plummeted 9% to $15.71 on Thursday afternoon — the stock’s biggest percentage decline in nearly a year — despite an initial bounce during the company’s inaugural Autonomy and AI Day event in Palo Alto.
The selloff came even as founder and CEO RJ Scaringe unveiled the company’s first custom silicon, the autonomy roadmap targeting Level 4 self-driving, and a driver-assistance subscription priced well below Tesla’s (more advanced Supervised FSD 14.2) offering.
The sharp decline suggests investors had priced in much of the optimism heading into the event.
Rivian shares had climbed more than 30% year-to-date before Thursday, driven by enthusiasm around the company’s technology roadmap and upcoming R2 SUV launch — scheduled for the first half of 2026.
Morgan Stanley had downgraded Rivian earlier this week to Underweight, citing concerns about the company’s ability to compete in the capital-intensive autonomous vehicle space, adding that ramping up the R2 and R3 models plays a more crucial role.
Autonomy+
Rivian announced its Autonomy+ driver-assistance subscription will launch in early 2026, priced at $2,500 as a one-time payment or $49.99 per month — significantly cheaper than Tesla‘s $8,000 upfront cost or $99 monthly subscription for Full Self-Driving.
The package will include Universal Hands-Free, a new capability bringing hands-free driving to more than 3.5 million miles of roads across the United States and Canada, covering the vast majority of marked roads.
Until March 2026, the software will be free to use.
‘Most Powerful’ Sensor Suite in NA
Vidya Rajagopalan, senior vice president of electrical hardware at Rivian, said the company expects that “at launch in late 2026, this will be the most powerful combination of sensors and inference compute in consumer vehicles in North America.”
The Gen 3 autonomy hardware centers on RAP1, Rivian’s proprietary 5nm processor that integrates processing and memory onto a single multi-chip module.
The Autonomy Compute Module 3 delivers 1,600 sparse INT8 TOPS and can process 5 billion pixels per second.
Rivian also confirmed LiDAR integration for future R2 models, providing three-dimensional spatial data and redundant sensing for edge-case detection.
Vertical Integration Strategy
The event showcased Rivian‘s vertically integrated approach spanning custom silicon, AI models, and software architecture.
“I couldn’t be more excited for the work our teams are driving in autonomy and AI,” Scaringe said. “Our updated hardware platform, which includes our in-house 1600 sparse TOPS inference chip, will enable us to achieve dramatic progress in self-driving to ultimately deliver on our goal of delivering L4.”
Scaringe described the technology as “an inflection point for the ownership experience — ultimately being able to give customers their time back when in the car.”
Large Driving Model and Data Flywheel
VP of Autonomy and AI James Philbin detailed Rivian‘s Large Driving Model (LDM), an LLM-like architecture that enables the company to apply advances in generative AI directly to its autonomy stack, including reinforcement learning.
The model is powered by a self-improving data flywheel that learns from the company’s deployed fleet, with pre-tagged instances transmitted to Rivian‘s cloud for automatic organization and labeling in near real-time.
Rivian Assistant Debuts
Chief Software Officer Wassym Bensaid introduced Rivian Assistant, a voice-activated AI companion that controls vehicle functions, manages calendars, sends messages, and works with third-party apps through an in-house agentic framework.
At R2 launch, a dedicated infotainment computer will enable Rivian Assistant to run entirely offline with higher performance and lower latency.
