Preferred Networks Raises $500M for Robotics AI

Preferred Networks, the Tokyo-based artificial intelligence startup backed by SoftBank Group, has raised approximately $500 million in fresh funding to supercharge its robotics AI ambitions. The massive funding round positions the company as one of Japan's most well-capitalized AI ventures and signals growing investor appetite for robotics-focused AI companies outside Silicon Valley.

The deal marks one of the largest AI funding rounds in Japanese history, rivaling investments typically reserved for U.S.-based heavyweights like OpenAI and Anthropic. It also underscores SoftBank CEO Masayoshi Son's deepening commitment to artificial intelligence as the cornerstone of his investment thesis.

Key Facts at a Glance

• Funding amount: Approximately $500 million in new capital
• Lead backer: SoftBank Group, with participation from existing and new investors
• Focus areas: Robotics AI, industrial automation, and deep learning infrastructure
• Valuation: Estimated to exceed $2 billion post-money, making it one of Japan's most valuable AI startups
• Founded: 2014, headquartered in Tokyo, Japan
• Core technology: Proprietary deep learning frameworks optimized for robotics and edge computing

SoftBank Doubles Down on Robotics AI

SoftBank's involvement in this round is no surprise. The Japanese conglomerate has been aggressively building an AI and robotics portfolio, from its acquisition of Boston Dynamics (later sold to Hyundai) to its ongoing investments in ARM Holdings and various AI startups through the Vision Fund.

Preferred Networks represents a strategic bet that is closer to home. Unlike many of SoftBank's global investments, Preferred Networks operates at the intersection of Japan's world-class manufacturing sector and cutting-edge AI research — a combination that few companies globally can replicate.

Masayoshi Son has repeatedly stated that AI and robotics will define the next era of technological advancement. This investment aligns with SoftBank's broader strategy of creating an integrated AI ecosystem, particularly as the company explores building AI-powered data centers and semiconductor supply chains.

What Makes Preferred Networks Different

Preferred Networks stands apart from typical AI startups in several critical ways. While companies like OpenAI and Google DeepMind focus primarily on large language models and generalist AI, Preferred Networks has carved out a niche in applied robotics intelligence — teaching machines to interact with the physical world.

The company's proprietary deep learning framework, Chainer, was once among the most popular open-source frameworks before the industry consolidated around PyTorch and TensorFlow. Although Preferred Networks eventually transitioned away from Chainer, the experience gave the team deep expertise in building AI infrastructure from the ground up.

Today, the startup focuses on several key verticals:

• Industrial robotics: AI systems that enable factory robots to handle complex, unstructured tasks
• Autonomous driving: Perception and planning algorithms for self-driving vehicles
• Drug discovery: Machine learning models for pharmaceutical research and molecular simulation
• Personal robotics: Consumer-grade robots for household tasks and eldercare
• Edge computing: Optimized AI inference on resource-constrained hardware

This diversified approach contrasts sharply with the 'foundation model or bust' mentality prevalent among Western AI startups. Preferred Networks builds specialized, vertically integrated solutions — an approach that may prove more commercially viable in the near term.

Japan's AI Ecosystem Gets a Major Boost

The $500 million raise sends a powerful signal about Japan's growing relevance in the global AI race. While the United States and China dominate headlines, Japan has quietly built a formidable AI ecosystem rooted in its manufacturing expertise and robotics heritage.

Japan faces a unique demographic challenge that makes robotics AI especially urgent. The country's population is aging rapidly, with over 29% of citizens aged 65 or older. Labor shortages in manufacturing, logistics, and healthcare are reaching critical levels, creating massive demand for intelligent automation.

Preferred Networks is not operating in a vacuum. Other Japanese AI companies, including Sakana AI (founded by former Google researchers) and various subsidiaries of Toyota, Fanuc, and Sony, are also investing heavily in AI-powered robotics. However, few have secured funding at this scale from a single backer as prominent as SoftBank.

The Japanese government has also been supportive, allocating billions of yen toward AI research infrastructure and semiconductor development as part of its national technology strategy.

The Global Robotics AI Race Heats Up

Preferred Networks' fundraise comes amid a broader surge of investment in robotics AI worldwide. In the United States, companies like Figure AI (backed by Microsoft and OpenAI), 1X Technologies, and Covariant have collectively raised billions to develop humanoid robots and intelligent automation systems.

Tesla's Optimus humanoid robot program and NVIDIA's Project GR00T robotics foundation model further illustrate the intensity of competition. Chinese competitors, including Unitree Robotics and various Tencent-backed ventures, are also racing to bring affordable robotic systems to market.

What distinguishes Preferred Networks in this crowded landscape is its decade-long track record of deploying AI in real-world industrial settings. While many robotics AI startups remain in the prototype phase, Preferred Networks has active partnerships with major Japanese manufacturers, including Toyota and FANUC, two of the world's largest industrial robotics companies.

This practical experience gives the startup a significant advantage in understanding the messy realities of deploying AI in factory environments — where reliability, safety, and precision matter far more than benchmark scores.

How the $500M Will Be Deployed

According to sources familiar with the company's plans, Preferred Networks intends to allocate the new capital across several strategic priorities. The bulk of the funding will go toward scaling its computing infrastructure, including acquiring large clusters of NVIDIA GPUs to train more sophisticated robotics models.

Additional investment areas include:

• Talent acquisition: Hiring top AI researchers and engineers from global talent pools
• International expansion: Establishing stronger commercial presence in the U.S. and European markets
• R&D acceleration: Developing next-generation foundation models specifically designed for robotic manipulation and spatial reasoning
• Partnership development: Deepening integrations with major manufacturers and logistics companies
• Hardware prototyping: Building proprietary robotic systems that showcase the company's AI capabilities

The international expansion component is particularly noteworthy. Preferred Networks has historically been Japan-centric, but the scale of this funding suggests ambitions that extend well beyond the domestic market.

What This Means for the Industry

For developers and engineers, Preferred Networks' raise validates the growing importance of robotics-specific AI skills. As the industry moves beyond chatbots and image generators, demand for professionals who understand both machine learning and physical systems engineering will surge.

For businesses, particularly in manufacturing and logistics, the investment signals that enterprise-grade robotics AI solutions are approaching commercial maturity. Companies that begin evaluating these technologies now will be better positioned to adopt them as they scale.

For the broader AI ecosystem, this funding round reinforces a critical trend: the next frontier of AI is not just digital — it is physical. Foundation models that can reason about and interact with the real world represent arguably the most consequential challenge in AI today.

Looking Ahead: Preferred Networks' Path Forward

The road ahead for Preferred Networks is both promising and challenging. The company must now justify its elevated valuation by demonstrating that its technology can scale beyond Japan's borders and compete with well-funded Western and Chinese rivals.

Key milestones to watch in the next 12 to 18 months include potential product launches in the U.S. or European markets, new strategic partnerships with global manufacturers, and the possible development of a robotics foundation model that could rival NVIDIA's offerings.

If Preferred Networks succeeds, it could become the most significant AI company to emerge from Japan since the current generative AI boom began. More importantly, it could prove that the future of AI extends far beyond large language models — into the factories, warehouses, and homes where intelligent machines will reshape daily life.

SoftBank's $500 million bet suggests that Masayoshi Son believes exactly that. The rest of the industry is watching closely.