XPeng is successfully licensing its self-driving technology to major automakers, a feat Tesla has failed to achieve despite five years of pitching its Full Self-Driving (FSD) system. The Chinese EV maker established a strategic commercialization team six months ago to shop its autonomous-driving software, smart cockpit systems, and AI chips to carmakers and suppliers worldwide. Meanwhile, Tesla's repeated attempts to license FSD to the industry have produced no signed deals, despite Elon Musk's public claims of ongoing negotiations.

What Makes XPeng's Licensing Deal Successful Where Tesla Failed?

The contrast between XPeng's traction and Tesla's empty hands comes down to practical business realities rather than technical superiority. XPeng already has a paying reference customer in Volkswagen, which invested roughly $700 million for a 4.99% stake in 2023 and is now mass-producing the ID.UNYX 08 electric SUV using XPeng's technology stack. That real-world deployment matters enormously to potential licensees evaluating risk.

XPeng's approach to autonomous driving mirrors Tesla's philosophy of using cameras as the primary driving system powered by end-to-end neural networks trained on massive amounts of data. However, XPeng retained a critical difference: redundant safety systems. The company kept radar and ultrasonic sensors as an independent safety layer that runs separately from the main driving AI, providing automatic emergency braking and steering as a backup even though cameras handle actual driving decisions.

For legacy automakers nervous about handing driving decisions entirely to artificial intelligence, that redundancy is far easier to underwrite and insure. Tesla deleted radar and ultrasonics in pursuit of pure vision-based driving, betting that cameras alone are sufficient. This design choice has become a sticking point in licensing negotiations, as traditional carmakers want the safety net that XPeng provides.

Why Has Tesla's FSD Licensing Strategy Repeatedly Stalled?

Tesla's licensing efforts have been remarkably consistent in their lack of results. Musk floated preliminary discussions with other carmakers in 2021, claimed Tesla was "happy to license" FSD after a 2023 charging deal with Ford, and asserted in April 2024 that Tesla was "in talks with one major automaker" about FSD. None of these discussions produced a signed agreement.

By November 2025, Musk acknowledged the reality publicly, writing on X that he had "tried to warn them and even offered to license Tesla FSD, but they don't want it! Crazy..." Ford CEO Jim Farley had already explained the quiet part out loud, telling the public that "Waymo is better." The liability question looms largest: legacy automakers reportedly demanded that Tesla accept the risk when a licensed system causes a crash, terms Musk called unworkable.

Musk

There is also a fundamental credibility gap. Tesla has spent years promising FSD to its own customers, yet the company still cannot point to an unsupervised product it has delivered to the millions of people who already paid for it. Potential licensing partners are watching this unfold and drawing their own conclusions about whether to trust Tesla with their brand's safety reputation.

How XPeng's Business Model Differs From Tesla's Approach

• Supplier Mindset: XPeng is willing to act as a technology supplier to other automakers, while Tesla has historically demanded platform economics and control, positioning itself as the standard everyone should adopt.
• Proven Deployment: XPeng has a mass-production reference customer in Volkswagen with cars already on the road, whereas Tesla's FSD remains a work-in-progress for its own vehicles.
• Safety Architecture: XPeng retained redundant sensor systems as an independent safety layer, making the technology easier for risk-averse legacy automakers to adopt and insure.
• Liability Stance: XPeng appears willing to negotiate liability terms that work for both parties, while Tesla has insisted on terms that shift crash risk entirely to the licensee.

XPeng's commercialization push is already showing financial results. Service revenue jumped 93.9% year-over-year to approximately $400 million in the second quarter of 2026, driven largely by the Volkswagen deal. This high-margin, non-hardware income is precisely what the licensing strategy is designed to scale.

What Does This Mean for the Broader Autonomous Driving Market?

The role reversal is striking: a Chinese EV startup is now the self-driving and electronics supplier to a legacy German automaker, while Tesla has spent half a decade insisting it would be the one to license FSD to the industry. XPeng's success suggests that automakers care less about which company has the most advanced AI and more about which company can offer a complete, insurable, deployable package with proven real-world results.

Geopolitical headwinds will likely limit XPeng's near-term licensing opportunities in the West. A US or European automaker licensing Chinese autonomous-driving software and Chinese AI chips faces significant regulatory and political obstacles in the current climate of data-security reviews and tariffs. The near-term buyers are probably other Chinese brands, emerging-market automakers, and suppliers rather than Detroit or Wolfsburg's rivals.

Still, the direction of travel matters. XPeng has proven that the licensing model works when executed with humility, redundancy, and a willingness to be a supplier rather than a platform dictator. Tesla's repeated failures suggest that technical prowess alone is not enough to win over risk-averse legacy automakers. The company's insistence on platform economics, combined with its inability to deliver a finished product to its own customers, has left the door open for competitors to capture the licensing opportunity Tesla once thought was its to lose.