In a sign of how venture capital interest in artificial intelligence and robotics continues to accelerate, Skild AI, a Pittsburgh-based startup specializing in advanced robotics foundation models, has closed a $300 million Series A funding round backed by some of the biggest names in global tech investment. The company’s ambitious plan to create “brain-like” AI systems for a wide range of robotic platforms has captured attention from investors betting on the future of automation, adaptive machines, and the next evolution of physical artificial intelligence.
Founded in 2023 by two former Carnegie Mellon University researchers, Skild AI has quickly made a name for itself with a focus on building general-purpose foundation models that aim to provide robots with the kind of adaptable intelligence humans take for granted. Rather than training robots on narrow, task-specific behaviours, Skild’s approach is to use massive datasets and cutting-edge AI techniques so that machines can learn, generalise and operate across a diversity of real-world environments — from warehouse logistics to manufacturing, inspection tasks and beyond.
The newly closed round values Skild AI at around $1.5 billion and was led by top tier venture firms including Lightspeed Venture Partners, Coatue, SoftBank Group and Bezos Expeditions, with participation from others such as Felicis Ventures, Sequoia Capital, Menlo Ventures, General Catalyst, CRV, Amazon’s Alexa Fund, SV Angel and Carnegie Mellon University itself. Such broad backing signifies strong investor confidence in the company’s vision and its potential to shape the future of embodied AI.
AI and robotics investment has been a standout theme so far in 2026 funding markets, with reports highlighting increasing deal activity across machine intelligence, infrastructure and robotics startups. As investors seek out transformative technology bets that can deliver growth over the next decade, companies like Skild are emerging as early leaders in what some analysts call the next frontier of AI — extending intelligence from data centres and screens into physical systems interacting directly with the world.
Skild AI’s foundation model concept is designed to be a shared “brain” for multiple robotic embodiments, meaning it could power the control systems of entirely different types of robots — whether mobile, humanoid, wheeled or specialised for industrial tasks. By training on data that captures complex physical interactions and environmental variability, these models aim to give robots a kind of adaptable intelligence far beyond traditional programmatic instructions. This could help machines understand context, infer actions and even transfer skills learned in one setting to tasks in another.
The significance of this approach lies in addressing one of the most persistent challenges in robotics: the generalisation gap. Historically, robots have excelled at narrowly defined tasks within controlled environments but struggled when faced with variability and unpredictability. Skild’s approach seeks to close that gap by leveraging advances in AI learning and large-scale data training — a method analogous to how modern generative models learn from diverse data to perform varied language or vision tasks.
Industry experts see Skild’s work as part of a broader movement to merge software intelligence with physical robotics, potentially accelerating automation across sectors such as logistics, construction, agriculture and even home services. Robots powered by adaptable AI could perform duties that today are either too dangerous or too costly for humans, while also supporting industries grappling with labour shortages.
However, observers note that technical and commercial challenges remain. Developing general-purpose robotic intelligence requires not just capital but also breakthroughs in simulation, perception, safety and the integration of AI with mechanical systems. Skild’s success will depend on continued progress in these areas, along with real-world field tests and partnerships that translate lab results into reliable commercial products.
For now, the $300 million injection of funding marks a pivotal milestone for Skild AI and places the company among a select group of robotics startups attracting substantial venture capital interest. As AI and robotics continue to converge, the race to build adaptable, intelligent machines capable of performing complex physical tasks is intensifying, and Skild’s foundation models could play a defining role in that story.
