Boston Dynamics and Google DeepMind announced a new artificial intelligence partnership at CES 2026 in Las Vegas to integrate Gemini Robotics AI foundation models into Boston Dynamics’ next-generation Atlas humanoid robots. The collaboration aims to enable humanoid robots to perform a wide range of industrial tasks, with initial applications focused on automotive manufacturing and joint research expected to begin this year. The companies describe the partnership as a step toward deploying adaptable humanoid robots across complex industrial environments.

Boston Dynamics has long been known for its robots’ athletic capabilities and dynamic movement. In 2024, the company formally announced its plan to build a commercial humanoid robot, citing recent advances in artificial intelligence that accelerated how robots can be trained and deployed in real-world environments.

The new partnership builds on that strategy.

Alberto Rodriguez, Director of Robot Behavior for Atlas at Boston Dynamics, said the company sought a partner capable of developing advanced visual-language-action models tailored to complex humanoid systems:

“We are building the world’s most capable humanoid, and we knew we needed a partner that could help us establish new kinds of visual-language-action models for these complex robots. Nobody in the world is better suited than DeepMind to build reliable, scalable models that can be deployed safely and efficiently across a wide variety of tasks and industries.”

Google DeepMind has been advancing robotics-focused AI foundation models in recent years, including Gemini Robotics, which is built on the large-scale, multimodal Gemini model. These models are designed to help robots perceive their environments, reason about tasks, use tools, and interact with humans.

Carolina Parada, Senior Director of Robotics at Google DeepMind, said:

“We developed our Gemini Robotics models to bring AI into the physical world. We are excited to begin working with the Boston Dynamics team to explore what’s possible with their new Atlas robot as we develop new models to expand the impact of robotics, and to scale robots safely and efficiently.”

The partnership centers on combining Boston Dynamics’ expertise in hardware, mobility, and robot control with DeepMind’s large-scale AI models.

The stated focus is industrial deployment, with automotive manufacturing identified as the initial application area. The companies describe the collaboration as a step toward enabling humanoid robots to perform a wide variety of industrial tasks with greater adaptability and reliability.

Joint research is expected to begin in the coming months and will take place across both organizations.

Q: What are Boston Dynamics and Google DeepMind building together?
A: The two companies are integrating Google DeepMind’s Gemini Robotics foundation models into Boston Dynamics’ next-generation Atlas humanoid robot to enable advanced industrial task execution.

Q: What is Gemini Robotics?
A: Gemini Robotics is a robotics-focused foundation model built on Google DeepMind’s multimodal Gemini AI system. It is designed to help robots perceive their environment, reason about tasks, use tools, and interact safely with humans.

Q: Why is Atlas central to this partnership?
A: Atlas is Boston Dynamics’ humanoid robot platform. The partnership aims to combine Atlas’ advanced physical mobility and control systems with DeepMind’s large-scale AI reasoning models.

Q: What industries are being targeted first?
A: The companies identified automotive manufacturing as the initial application area for humanoid deployment.

Q: When does the joint research begin?
A: Research is expected to begin this year and will be conducted collaboratively at both organizations.

This partnership represents more than a technical collaboration—it reflects the growing convergence of large-scale AI foundation models and embodied humanoid robotics in industrial environments.

Who should care:
Manufacturing leaders, automotive executives, robotics engineers, and plant operations teams evaluating advanced automation will want to track this development closely. For companies operating complex production lines, the evolution of adaptable humanoid systems could influence how future facilities are designed and staffed.

Why it matters now:
Multimodal AI foundation models are no longer confined to software interfaces or cloud-based assistants. They are beginning to interface directly with physical systems. If Gemini Robotics can reliably translate visual-language reasoning into real-world motor control inside Atlas, it could expand the scope of tasks humanoid robots are able to perform—particularly in environments that are too variable or complex for traditional fixed automation.

That represents a meaningful technical threshold: the possibility of robots adapting within industrial spaces rather than executing only pre-programmed routines.

What decision this affects:
For industrial and automotive manufacturers, this development may influence long-term decisions around capital investment, workforce planning, plant flexibility, and the balance between specialized machinery and general-purpose robotic systems. It also raises practical considerations around safety protocols, system integration, and return-on-investment timelines for next-generation automation platforms.

While this announcement outlines a research collaboration rather than immediate deployment, it marks a concrete step toward testing advanced AI reasoning models inside commercially oriented humanoid hardware. The results of that experimentation will determine how quickly humanoid robotics moves from demonstration stages into sustained industrial use.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.