Skild AI’s $1.4 billion funding round, led by SoftBank, just 18 months after launch, validates that robotics foundation models will dominate industrial automation, with the economics of physical AI and robot deployment shifting decisively toward software platforms rather than hardware manufacturing.

The Pittsburgh-based startup, valued at $14 billion, secured funding from NVIDIA Ventures, Samsung, LG, and Jeff Bezos, tripling its valuation to $14 billion from $4.5 billion in May 2025. The financing signals institutional conviction that foundation models controlling diverse robot embodiments will capture more value than vertical hardware plays. This strategic inflection redirects capital flows across the robotics AI sector.

Skild AI is developing the Skild Brain, the industry’s first unified robotics foundation model designed to control any robot regardless of its physical form. Unlike traditional models built for specific machines, the omni-bodied Skild Brain can operate quadrupeds, humanoids, robotic arms, and mobile manipulators, enabling tasks that range from everyday household chores to complex physical navigation.

To overcome the lack of large-scale robotics data, Skild AI trains the model by learning from human videos on the internet and through physics-based simulations, rather than relying on narrowly deployed, task-specific robots. This approach allows the Skild Brain to generalize across different robot morphologies and environments.

Crucially, the model can adapt to unpredictable conditions such as damaged limbs, mechanical failures, heavier loads, or even entirely new robot bodies, without retraining, positioning Skild AI’s technology as a major step toward truly general-purpose robotics.

The Data Bottleneck Solution

Founded in 2023 by Carnegie Mellon professors Deepak Pathak and Abhinav Gupta, Skild AI is building a scalable, general-purpose robotics foundation model that serves as a shared brain across diverse robot forms and real-world tasks.

With offices in Pittsburgh, the San Francisco Bay Area, and Bengaluru, the company aims to overcome robotics’ core challenge of data scarcity by commercializing advances in self-supervised and adaptive robotics developed by its founders.

“Unlike language or vision, there is no robotics data available on the internet. Therefore, you cannot simply apply these generative AI methods,” Deepak Pathak told Reuters in July 2025, explaining how Skild AI trained its foundation model on 100,000 simulated robots over one million simulated years, augmented by learning from millions of human action videos.

This approach unlocked in-context learning for physical systems, enabling the Skild Brain to adapt in real time to hardware failures, missing limbs, or entirely new robot bodies without retraining.

“We believe this omni-bodied learning is essential for building AGI that works reliably in the physical world, paving the way for robots that can safely help humans in everyday environments,” said Abhinav Gupta, Co-Founder and President of Skild AI. “This enables robots to operate dynamically in complex environments, without requiring preprogrammed instructions for each scenario.”

The company claims robot deployment costs drop to $4,000-$15,000 per unit compared to $250,000 for traditionally programmed systems, creating discontinuous economics that unlock previously unviable automation use cases across industrial automation.

SoftBank’s Vertical Integration Play

SoftBank’s lead investment directly complements its October 2025 acquisition of ABB Robotics for $5.4 billion, orchestrating control of both physical hardware and adaptive intelligence. ABB brought $2.3 billion in annual revenue and four million installed industrial robots; Skild AI delivers the software layer that CEO Masayoshi Son believes defines the robotics foundation that models the future.

If Skild AI’s robotics foundation models scale across ABB’s installed base, SoftBank captures recurring software revenue on established hardware infrastructure. However, integration carries execution risk; SoftBank’s previous robotics venture, the Pepper humanoid, failed commercially despite comparable ambitions.

“Skild AI is building foundational technology for Physical AI across robots, tasks, and environments,” said Dennis Chang, Managing Partner at SoftBank Investment Advisers. “We’re proud to partner with Deepak, Abhinav, and the Skild AI team to bring that shared vision into real-world applications worldwide.”

Commercial Traction Signals Market Inflection

Skild AI reached approximately $30 million in annualized revenue within months of commercializing in 2025, exceptionally rapid for foundation model companies. Current deployments include LG CNS partnerships for manufacturing automation, LaGuardia Airport monitoring robots, and pilots across eight enterprise customers.

LG CNS CEO Hyun Shin-gyoon predicted at CES 2026 that physical AI would reach commercial deployment within two years. “While we are already conducting proof-of-concept projects using robots in manufacturing settings, additional time is needed to prepare the broader environment,” Hyun said, noting tests at more than 10 customer facilities.

This timeline aligns with industry consensus that 2025-2026 marks the inflection point at which pilots transition to production scale. The global AI robots market is projected to expand from $20.5 billion (2025) to $124.26 billion (2034), with a 22.16 percent CAGR, driven by labor shortages and improving robotics foundation models.

Platform Economics vs. Hardware Integration

Skild AI’s funding validates strategic divergence, reshaping investment. Traditional companies pursued vertical integration, while Boston Dynamics optimized hardware and software for specific forms. Figure AI, valued at $39.5 billion, manufactures humanoid robots with integrated intelligence.

Skild AI rejects this approach, positioning itself as horizontal infrastructure analogous to cloud platforms. The company operates with near-zero marginal cost for software deployment, enabling rapid scaling without manufacturing constraints. Strategic investors, including Samsung, LG, Schneider Electric, and Salesforce, signal adoption bets across manufacturing, healthcare, and enterprise IT.

“Solving intelligence for the physical world unlocks enormous commercial value and long-term strategic national importance,” said Rita Waite, Partner at IQT. “Skild AI is uniquely positioned to do both.”

Yet execution risk remains material. An estimated 98 percent of robot deployments in manufacturing fail to meet objectives due to limitations in adaptability. Robotics foundation models must demonstrate generalization across diverse real-world environments, not just in laboratory demonstrations, to justify extreme valuations.

Competitors, including Physical Intelligence (raised $400 million, valued at $5.6 billion) and Tesla’s Optimus, pursue similar strategies, fragmenting the market. Concerns about an “AI winter” loom—Professor Michael Fisher warned in August 2025 that “over-promising the benefits of AI is likely to lead to a collapse in confidence,” potentially stalling development if bubbles burst.

Network Effects and Value Capture

The $14 billion valuation assumes Skild AI establishes defensible network effects: each robot deployment generates real-world data that improves the shared foundation model and creates continuous feedback loops. As models improve, robot deployment costs decrease and capabilities expand, attracting more customers and accelerating the data flywheel.

This dynamic mirrors the advantages of concentrated value in cloud computing, where a handful of winners dominate. However, physical AI faces unique constraints; robot deployment requires hardware procurement, installation, training, and maintenance, which slows adoption cycles and limits winner-take-all dynamics.

For manufacturers confronting labor shortages, the US faces 2.1 million unfilled manufacturing jobs by 2030. Skild AI’s economics potentially unlock industrial automation previously considered unviable. Yet the strategic question remains, will horizontal platforms or vertical integration capture more value as robotics matures?

Skild AI’s $1.4 billion represents a decisive bet that software generalization outpaces hardware specialization. SoftBank’s simultaneous control of ABB’s manufacturing infrastructure and Skild AI’s intelligence layer hedges this uncertainty, positioning to capture value regardless of which model dominates.

For industrial automation broadly, the funding signals robotics foundation models have transitioned from research to infrastructure investment, with capital abundance accelerating production deployment timelines to 24 months.

Follow USTechTimes on Facebook, Twitter and Linkedin for in-depth news of market trends, funding updates, and regulatory changes affecting startups in USA.

We Recommend: