For two years, AI has been essentially a screen-based phenomenon. It wrote your emails, generated bizarre images, summarized documents you didn't want to read, and answered questions with varying degrees of accuracy. Useful, sure. But ultimately confined to pixels and text boxes.

That era is ending.

We're entering what Bank of America calls Physical AI: intelligence that doesn't just process information but actually interacts with the physical world. Robots that can assemble cars. Vehicles that drive themselves through city traffic. Drones that deliver packages to your doorstep. AI isn't just thinking anymore. It's doing.

Bank of America's thematic investing team, led by analyst Martyn Briggs, is calling this a trillion-dollar transformation that's already in motion. The value is migrating from digital models to physical machines that can see, reason, and act autonomously in real environments.

"Intelligence moves from screens to machines," Briggs wrote in a report released Thursday.

Nvidia CEO Jensen Huang is even more direct about where we are in this transition.

"The ChatGPT moment for physical AI is here — when machines begin to understand, reason, and act in the real world. The next wave of AI is physical AI," Huang declared at CES 2026.

Understanding the Physical AI Revolution

So what exactly separates Physical AI from the generative AI tools we've been using? It's the difference between a system that can describe how to change a tire and one that can actually change the tire.

Bank of America defines Physical AI as systems embedded in machines like humanoid robots, autonomous vehicles, and drones that can perceive their environments, reason about what they're sensing, and execute physical actions. These systems rely on multimodal world models trained with both visual and action data, enabling them to predict outcomes, plan sequences, and operate without constant human supervision.

The technology is already moving beyond research labs. Robotaxis are operating commercially in multiple cities right now. In China, advanced driver-assistance systems are on track for mass adoption by 2030.

As hardware costs decline and onboard computing power increases, the economics are starting to work. Removing human drivers from ride-hailing and freight transportation could dramatically cut costs, essentially transforming mobility into a software-defined, AI-driven platform rather than a labor-intensive service.

Humanoid Robots Are Actually Happening

"Humanoid robots have emerged as the most visible frontier of this shift," Briggs noted.

More than 50 companies are now developing humanoid robot platforms. Early deployments are concentrated in manufacturing, logistics, and hazardous environments where the robots can take on dangerous or repetitive tasks.

The trajectory is striking: shipments are expected to grow from tens of thousands of units today to millions annually within the next decade. The drivers are clear enough: persistent labor shortages in key industries, falling costs for components like actuators and sensors, and vertical integration strategies that let companies control more of the technology stack.

The companies that win this market won't just be good at building robot bodies. They'll need to excel at actuators, sensors, energy efficiency, and the AI software that integrates everything into a functioning autonomous system.

The 15 Companies Leading the Physical AI Charge

Bank of America identified several publicly traded companies with significant exposure to Physical AI across semiconductors, robotics, mobility, and sensing technologies. Here's the full list of 15 stocks analysts believe are positioned to lead this transformation.

1. Nvidia Corp. (NVDA)

The undisputed infrastructure leader in Physical AI, Nvidia offers a full-stack platform spanning robotics, mobility, and industrial applications. At CES 2026, the company launched GR00T and Cosmos, technologies designed to enable autonomous reasoning across different types of machines.

2. Tesla Inc. (TSLA)

Tesla now operates robotaxis in nine cities and is scaling both its autonomous vehicle fleet and licensing arrangements for its Full Self-Driving software. Weekly ride volumes tripled throughout 2025.

3. Qualcomm Inc. (QCOM)

Qualcomm's Dragonwing IQ10 system-on-chip powers physical AI at the edge, enabling low-power decision-making for drones, robots, and autonomous vehicles that need to operate beyond cloud connectivity.

4. Advanced Micro Devices Inc. (AMD)

AMD's open AI ecosystem (ROCm) and scalable hardware position it as a key alternative to Nvidia, targeting both cloud training infrastructure and edge inference markets.

5. Ambarella Inc. (AMBA)

A pioneer in power-efficient AI video chips, Ambarella demonstrated AI-optimized SoCs at CES capable of real-time vision and reasoning in constrained environments.

6. ARM Holdings plc (ARM)

With embedded chips in everything from smartphones to robots, ARM enables physical AI to run directly on devices, reducing latency and power consumption.

7. Hitachi Ltd. (HTHIY)

Hitachi's HMAX AI stack integrates operational and sensor data to improve uptime, control, and prediction capabilities across mobility, infrastructure, and cybersecurity applications.

8. Innoviz Technologies Ltd. (INVZ)

This LiDAR supplier is benefiting from an over 99% cost decline in sensor technology since 2019. Innoviz is helping expand autonomous vehicles and robotics beyond research and development into actual commercial deployment.

9. BYD Co. Ltd. (BYDDF)

China's smart electric vehicle leader is seeing surging ADAS penetration, with "God's Eye" navigation technology deployed in high-end models and L2+/L3 autonomous systems scaling rapidly.

10. Alphabet Inc. (GOOGL)

Waymo's robotaxi operations are scaling quickly, with hardware costs down 50–70%. Commercial activity is intensifying across multiple U.S. urban markets.

11. Baidu Inc. (BIDU)

Baidu's Apollo fleet is part of China's ambitious roadmap to deploy 400,000 robotaxis by 2030. The company is positioned as a dominant autonomous vehicle software stack provider in Asia.

12. AeroVironment Inc. (AVAV)

AeroVironment's unmanned aircraft systems dominate tactical drone markets. Defense and public security demand is accelerating, with drones deployed in both surveillance and disruption roles.

13. Kratos Defense & Security Solutions Inc. (KTOS)

Specializing in autonomous defense technology and simulation, Kratos offers high-performance, low-cost systems critical for modern military AI operations.

14. Apple Inc. (AAPL)

Still a dark horse in this space, Apple's on-device AI chips, spatial computing stack, and long-rumored Project Titan suggest ambitions that extend well beyond smartphones.

15. Li Auto Inc. (LI)

This leading Chinese automaker is rapidly integrating advanced L2+ autonomy features and is expected to be part of China's broader autonomous vehicle expansion.

The shift from digital AI to Physical AI represents more than just a technological evolution. It's a fundamental change in how artificial intelligence creates value, moving from productivity tools on screens to autonomous systems that can navigate and manipulate the physical world. Bank of America's list suggests the investment opportunities span the entire stack, from the chips that power these systems to the platforms that deploy them at scale.