YY Group Pivots to AI Training Data With 500,000 Users

YY Group is betting its 500,000-person workforce can be transformed into a high-fidelity data source for the booming artificial intelligence industry.

YY Group Holding Limited (Nasdaq: YYGH), a Singapore-based workforce solutions provider, is strategically pivoting into the AI infrastructure market with a new platform to generate real-world training data for robotics and AI systems. The move aims to address a critical bottleneck in AI development: the scarcity of structured data on human physical tasks, which is essential for training autonomous systems.

"This strategic initiative reflects the evolution of YY Group's platform capabilities beyond workforce services and into the AI data infrastructure layer," said Mike Fu, Chief Executive Officer of YY Group, in a statement. "By leveraging our global network, we are uniquely positioned to generate the real-world human data required for AI systems to operate effectively in physical spaces."

The company is establishing a dedicated data collection facility in Johor, Malaysia, to capture human workflows in service environments. These operational tasks will be converted into datasets for training AI models and humanoid robots. The entire operation will be powered by YY Group's YY Circle App, which can mobilize its network of more than 500,000 users across 12 countries into paid "Data Collection" shifts.

The pivot is designed to unlock higher-margin revenue streams and position YY Group as a key data partner for global technology companies. By converting its existing labor force into a data-generation asset, the company is entering the lucrative AI supply chain, a departure from its traditional business of on-demand staffing and facilities management.

The Race for Real-World Data

YY Group's strategy comes as the world's largest technology firms scramble for new sources of high-quality training data, the lifeblood of all AI models. The move echoes efforts inside major labs, such as Meta Platforms Inc., which recently began using internal employee keystrokes and mouse movements to train its own AI models, according to reports. The shared goal is to capture how humans actually interact with digital and physical systems to make AI agents more effective.

While tech giants like Google are developing specialized hardware like Tensor Processing Units (TPUs) to handle massive training and inference workloads, the underlying models still depend on vast and varied datasets. YY Group is betting that its access to a large, globally distributed human workforce can provide a unique and scalable solution to the data-gathering problem, particularly for tasks that occur in the physical world, not just on a computer screen.

From Workforce to AI Infrastructure

The transition from a workforce solutions company to an AI data provider is not without precedent, though it carries risks. The market has recently seen speculative pivots, such as footwear company Allbirds rebranding to NewBird AI to lease GPUs, a move that was met with initial investor excitement followed by a sharp correction.

However, YY Group's approach differs significantly. Instead of simply acquiring and leasing hardware, the company is integrating its core asset—a massive, organized labor network—into the AI value chain. This strategy aims to create a defensible moat by providing something hardware alone cannot: structured, real-world human activity data at scale. The success of this pivot will depend on the company's ability to execute on its data collection plans and secure partnerships with the major AI and robotics developers who are the ultimate customers for this new asset class.

This article is for informational purposes only and does not constitute investment advice.