2026-06-18 👁️ — views

Physical AI Platform Licensing 2026 — Waymo Hyundai OEM Deal vs Tesla Closed Vertical Stack: The AV Business Model Benchmark

Waymo's Hyundai deal puts factory-integrated AV sensors in an OEM platform. Tesla licenses nothing and keeps its closed stack. Two opposite AV business models.

Article 193 in the Physical AI Benchmark Series — Physical AI Platform Licensing 2026: Waymo’s Hyundai and Moove Partnerships vs Tesla’s Closed Vertical Stack — The AV Business Model Benchmark

One of the least-discussed dimensions of the Waymo vs Tesla Physical AI race is not about software maturity, sensor count, or miles driven. It is about business model architecture: who owns the supply chain, who manufactures the vehicles, and who captures the economic value when autonomous driving works at scale. Waymo is building toward an open platform that can license its autonomous driving technology to other automakers and fleet operators. Tesla keeps its entire stack — hardware, software, data, manufacturing — vertically integrated and proprietary. This difference in business model architecture determines who can achieve global AV scale faster, and at what capital cost.

The Waymo-Hyundai partnership for the Ioniq 5 Robotaxi is the clearest signal that the open platform AV model is gaining commercial traction. A global OEM building factory-integrated Waymo hardware is a qualitatively different relationship than retrofitting a production vehicle — it is the beginning of a model where AV software companies and automotive manufacturers specialize in what they do best. Tesla’s closed stack creates integration depth that the platform model cannot match, but limits the technology’s spread to Tesla’s own production capacity. Understanding these two models is essential for assessing which company can achieve global AV deployment at scale.

Section 1 — Open Platform vs Closed Stack: The Fundamental AV Business Model Choice

The AV industry has converged on two dominant business model archetypes, and every major player must choose between them — or attempt to operate both simultaneously.

Archetype 1: Open Platform / Technology Licensor

The platform model separates the development of autonomous driving software and sensor technology from the manufacturing and operation of vehicles. The AV technology company develops the software stack, the sensor suite, and the integration know-how. OEMs manufacture vehicles incorporating the AV technology. Fleet operators deploy and operate those vehicles as commercial services. The technology company earns revenue through licensing fees, hardware supply agreements, and potentially software royalties — without bearing the capital cost of every vehicle in service.

Waymo is the clearest example of a company evolving toward this model, though it remains a hybrid: currently operating its own fleet AND pursuing partnerships that expand the technology’s reach through OEM and fleet operator channels.

Archetype 2: Closed Vertical Stack

The closed model integrates vehicle design, manufacturing, AI software development, training data infrastructure, and fleet operation under a single company. Tesla exemplifies this approach: Tesla designs and manufactures the vehicle, develops FSD in-house, retains all training data within its ecosystem, and plans to operate Robotaxi service under the Tesla brand. No component of the technology stack is licensed to competitors or partners.

Why the choice matters enormously for scale

The platform model allows AV technology to spread through partners’ manufacturing and distribution networks without the technology company bearing all capital costs. Each OEM partner that adopts the platform multiplies the number of vehicles carrying the AV system without a corresponding increase in the platform company’s capital expenditure. This is the Android model: Google developed the operating system and licensed it to OEMs, achieving 70%+ global smartphone market share through partner manufacturing — a scale that Apple’s closed iOS model could not match on volume, even if it dominated on margin.

The historical analogies are directionally important even if imperfect:

The AV analogy is imperfect in important ways: autonomous driving is more safety-critical than a smartphone OS, the integration between hardware and software is tighter, and the liability structure is different. But the directional lesson holds — a platform that achieves wide OEM adoption has structural scale advantages over a closed model constrained by one company’s manufacturing capacity.

The key conflict in Waymo’s two-track strategy

Waymo currently operates its own commercial fleet (the operator role) while also pursuing OEM partnerships that could eventually support external licensing (the platform role). These roles can conflict: if Waymo licenses the Waymo Driver to Hyundai for Hyundai’s own non-Waymo-operated fleet, Waymo is simultaneously a technology supplier and a competitor to that fleet in the same geographic market. Managing that conflict — and deciding when to prioritize operator revenue vs licensing reach — is the central strategic tension in Waymo’s business model evolution.

Tesla has no such tension. Its one-track approach controls everything from cell chemistry to FSD software to planned Robotaxi operations. This creates integration depth advantages but a hard ceiling on scale tied to Tesla’s own manufacturing capacity — unless the Tesla Network peer-to-peer model unlocks owner-deployed vehicles at scale.

Section 2 — Waymo’s Partnership Portfolio

Waymo has assembled a multi-layer partnership structure that spans vehicle manufacturing, fleet operations, and ride distribution. Each partnership expands Waymo’s reach through a different mechanism.

The Hyundai partnership deserves special attention as the most significant OEM deal in AV history. Prior Waymo vehicle partnerships — the Chrysler Pacifica, the Jaguar I-PACE — involved retrofitting production vehicles with Waymo hardware after the vehicle left the factory. The Ioniq 5 Robotaxi is being built differently: Hyundai integrates Waymo’s sensor hardware at the factory level, meaning the vehicle is designed and manufactured with AV capability as a first-order requirement rather than an afterthought. This is the factory-integration model that the AV industry has theorized about for years but that has not been executed at commercial scale until now.

The Moove partnership illustrates a different dimension of the platform strategy: fleet operations expertise as a partnership asset. Waymo cannot build local fleet operations knowledge in every new geographic market from scratch — it would require years of local presence, regulatory relationships, and operational infrastructure. Moove already has this across multiple markets. By partnering with Moove, Waymo can enter new markets faster and at lower capital cost by leveraging Moove’s existing operational footprint.

Section 3 — Tesla’s Closed Vertical Stack

Tesla’s approach to autonomous vehicles is the inverse of Waymo’s partnership model at every dimension. Tesla designs and controls every layer of its AV stack.

The closed vertical stack creates an integration depth that is genuinely difficult to replicate. FSD is not just software running on generic hardware — it is software co-optimized for Tesla’s specific camera placement, resolution, and frame rate. The Dojo training cluster is purpose-built to process the specific data format that Tesla’s cameras generate. The inference chip in every Tesla vehicle is designed to run Tesla’s specific neural network architecture efficiently. These integrations compound: each layer of the stack that is co-optimized with every other layer creates a system that performs better than the sum of independently designed parts.

The data flywheel is the most powerful aspect of the closed stack. Tesla’s 6M+ vehicle fleet generates training data at a scale that no other AV company can approach. Every edge case that a Tesla encounters — unusual lane markings, construction zones, aggressive merging behavior — becomes training data that improves FSD for every other Tesla on the road. This closed loop between fleet data and model improvement creates a competitive moat that is structural, not just technical: even if a competitor built software as good as FSD today, they would need years of fleet deployment to generate the training data necessary to maintain parity as conditions evolve.

The limitation is equally structural: Tesla’s AV technology can only spread as fast as Tesla can manufacture vehicles. Gigafactory capacity is a hard constraint on how quickly FSD and future Robotaxi technology can reach new users. No partner OEM will carry the technology. No fleet operator will operate non-Tesla vehicles under Tesla’s AV umbrella.

Section 4 — Platform vs Stack: The Business Model Economics

The economic implications of the two business models become most visible when analyzed at the dimensions that determine long-term profitability and scalability.

The profitability timeline comparison is the most practically important near-term economic dimension. Tesla already has a revenue stream from FSD subscriptions and hardware upgrades — the AV technology generates revenue today, even if autonomous ride services are not yet at scale. Waymo is operating at a loss and depends on Alphabet funding to sustain its fleet operations while the commercial model develops. This asymmetry means Tesla can sustain its AV investment from its own revenue base in a way that Waymo currently cannot — an important durability consideration if the timeline to AV profitability extends longer than expected.

The long-term economic comparison inverts if Waymo’s licensing model executes. Software licensing revenue — technology royalties on millions of partner vehicles, without the capex of owning those vehicles — is the highest-margin business model in technology. If Waymo achieves the Android-of-AV scenario, it collects licensing fees on every vehicle in every OEM partner’s fleet at near-zero marginal cost per additional vehicle. Tesla’s vertical model cannot approach that margin structure because Tesla must build and sell every vehicle that carries its AV technology.

Section 5 — Business Model Benchmark Scorecard

Overall verdict: The Waymo-Hyundai partnership is the clearest signal that the open platform AV model is gaining commercial traction. A global OEM building factory-integrated Waymo hardware means the cost and quality of Waymo’s fleet will improve through Hyundai’s manufacturing scale — the partnership-as-cost-reduction model that distinguishes platform strategy from pure operator strategy. Tesla’s closed vertical stack creates unmatched integration depth, a structural training data advantage, and a unique peer-to-peer Network potential via its existing owner fleet — but limits technology spread to Tesla’s own production capacity.

The long-term question is whether the AV industry follows the smartphone model (iOS closed vs Android open, with Android winning by volume while Apple captured profit share) or the luxury vehicle model (closed integration wins on quality premium). The answer may depend on whether autonomous driving is a commodity feature (favors open platform) or a differentiated safety/experience product (favors closed integration). If autonomous driving becomes a commodity where any competent system is good enough, Waymo’s platform model wins by volume. If it remains a differentiated product where integration quality matters enormously for safety and passenger experience, Tesla’s closed stack wins on depth.

The business model choice is not reversible quickly. Tesla cannot open-source FSD and build a partner ecosystem overnight. Waymo cannot build Tesla’s data flywheel without a consumer fleet. Both companies have locked in structural bets that will play out over the next five to ten years — and the Hyundai partnership is the strongest evidence yet that Waymo’s bet is finding commercial validation.

Note: Figures labeled “(est.)” are directional estimates based on publicly available information as of mid-2026. Partnership terms, licensing discussions, and business model plans not fully publicly disclosed are described based on publicly available statements and industry analysis. This article does not constitute investment advice.

Sources

• Waymo and Hyundai partnership for Ioniq 5 robotaxi — Waymo blog ↗
• Waymo and Uber partnership — Waymo press room ↗
• Tesla FSD and the closed vertical stack — Tesla AI Day ↗
• Waymo and Moove fleet operations partnership — Moove press ↗

Tags