Tesla’s ambitious push into artificial intelligence is advancing on two distinct fronts. The company is aggressively building its own semiconductor infrastructure while remaining heavily reliant on external hardware from Nvidia. This technological drive, however, is encountering headwinds from U.S. regulators, casting a shadow over the company’s expansive ambitions.
Regulatory Pressure Intensifies
Beyond its technological advances, Tesla’s software for autonomous driving is once again under regulatory scrutiny. The U.S. National Highway Traffic Safety Administration (NHTSA) escalated its probe into the “Full Self-Driving” system to an engineering analysis on Thursday. This investigation, now covering approximately 3.2 million vehicles, was prompted by incidents occurring under poor visibility conditions.
The market has reacted cautiously to this confluence of events. Shares declined by 3.23 percent to 332.15 euros today, extending the year-to-date loss to more than eleven percent.
Despite the regulatory pressure and weak share performance, analysts remain deeply divided in their assessment of Tesla’s transition from a pure automaker to an AI and robotics enterprise. This fundamental discrepancy is reflected in the wide range of current price targets, from UBS’s $352 to Bank of America’s $460.
Building an In-House Semiconductor Foundation
To meet the immense computational demands of autonomous driving and humanoid robots, Tesla is breaking ground on a specialized facility for AI chip production on March 21. Dubbed the “Terafab Project,” this initiative marks a shift toward becoming an integrated component manufacturer. The internally developed AI5 chips are designed to operate directly within vehicles and the future robotaxi fleet. Elon Musk claims that targeted software optimization will deliver performance significantly exceeding the chips’ raw specifications. For the computationally intensive model training in data centers, however, Nvidia’s graphics processing units remain essential.
The company’s vision already extends to the next hardware generation. Reports indicate Tesla is collaborating with Samsung Electronics on a multi-billion dollar deal to develop the planned AI6 chip. The ambitious timeline outlines the following milestones:
- December 2026: Completion of chip design (tape-out)
- Second Half of 2027: Commencement of mass production using a 2-nanometer process
- From 2028 Onward: Planned integration into vehicles and Optimus robots
This dual-track approach—combining immediate external sourcing with long-term internal development—defines Tesla’s strategy to secure its position at the forefront of AI hardware.