Apple's top multimodal AI researcher has joined autonomous vehicle company Wayve to lead a new robotics division, suggesting that foundation models developed for self-driving cars can transfer to broader physical AI applications. Alex Toshev, who previously led multimodal foundation model research at Apple, has been appointed Research Director at Wayve Labs to establish a dedicated robotics team focused on robot manipulation beyond automotive use cases.

Who Is Alex Toshev and What's His Track Record?

Toshev brings two decades of AI and robotics expertise to the role. At Apple, he co-led the development of MM1, the company's multimodal large language model family, which combines text and visual understanding in a single AI system. He also led research on generalist agents, AI systems that can interact with applications through graphical user interfaces without task-specific training.

Before joining Apple, Toshev spent more than a decade at Google, including six years in Google Robotics. There, he co-led SayCan, a landmark project that introduced large language models into robotic planning and won the Best Innovation Paper award at CoRL 2022, a top robotics conference. This combination of foundation model expertise and proven robotics deployment experience makes him a significant hire for Wayve's expansion beyond autonomous vehicles.

Why Is Wayve Betting on Foundation Models for Robotics?

Wayve's decision to hire Toshev and establish a separate robotics team signals a strategic shift in how the company views its core technology. Rather than positioning itself solely as an autonomous driving specialist, Wayve is framing itself as a foundation model company whose underlying technology can apply to multiple physical embodiments, including robots. Autonomous vehicles, in this view, become an early proving ground rather than the end goal.

The new robotics team, starting with a founding group in Wayve's Sunnyvale office, will combine Wayve's foundation model platform and infrastructure with robotics-specific research. The work will span hardware and software development, data collection, model training, evaluation, and deployment across different physical systems.

"Toshev brought exceptional expertise across computer vision, machine learning and robotics to build a new effort at the intersection of robotics and foundation models," said Jamie Shotton, Chief Scientist at Wayve.

Jamie Shotton, Chief Scientist at Wayve

How Does This Hire Accelerate Wayve's Robotics Ambitions?

• Validated Deployment Experience: Toshev's work on SayCan and generalist agents at Google Robotics provides Wayve with research already proven at major technology companies, potentially accelerating a robotics effort that would otherwise take years to build from scratch.
• Multimodal AI Expertise: His leadership of MM1 at Apple demonstrates deep knowledge of combining multiple types of data (text, images, video) into unified AI systems, a capability essential for robots that must understand and interact with complex physical environments.
• Separation from Commercial Pressures: Establishing a dedicated Sunnyvale-based robotics team keeps this research distinct from Wayve's automotive commercial priorities, allowing longer-term foundational research without competing directly against near-term deployment demands from partners like Nissan, Stellantis, and Uber.

Wayve Labs itself was launched in June 2026 as the company's frontier research unit, with major automotive partners including Nissan, Stellantis, and Uber already committed to deploying Wayve's autonomous driving technology. The addition of a robotics division suggests Wayve intends to leverage the same foundation model architecture across multiple industries and applications.

What Does This Mean for the Broader AI Industry?

The hire reflects a broader trend in AI development: companies that build powerful foundation models are increasingly exploring how those models can transfer across different domains and physical systems. If Wayve's approach succeeds, it could demonstrate that the same AI architecture used to train autonomous vehicles can be adapted for warehouse robots, manufacturing systems, and other physical automation tasks. This would represent a significant shift from the current model, where robotics and autonomous driving are typically developed as separate specialties with different technical approaches.

The move also underscores the competitive importance of recruiting top AI talent from established tech companies. By bringing Toshev from Apple, Wayve gains not just an individual researcher but access to validated techniques and institutional knowledge from one of the world's largest AI research organizations.